Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Contents lists available at ScienceDirect

Progress in Nuclear Magnetic Resonance Spectroscopy journal homepage: www.elsevier.com/locate/pnmrs

Solid-state NMR of quadrupolar halogen nuclei Rebecca P. Chapman, Cory M. Widdifield, David L. Bryce * Department of Chemistry and Centre for Catalysis Research and Innovation, University of Ottawa, Ottawa, Ont., Canada K1N6N5

a r t i c l e

i n f o

Article history: Received 31 March 2009 Accepted 5 May 2009 Available online 15 May 2009

Crown Copyright Ó 2009 Published by Elsevier B.V. All rights reserved.

Keywords: Chlorine Bromine Iodine Chemical shift Quadrupolar coupling constant Nuclear magnetic resonance NMR Tensors

Contents 1. 2.

3.

4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theory, conventions, and chemical shift referencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Magnetic shielding interaction and chemical shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1. Haeberlen–Mehring–Spiess (HMS) convention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2. Herzfeld–Berger/Maryland convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Nuclear electric quadrupolar interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. NMR properties of chlorine, bromine, and iodine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental methods for powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Magic-angle spinning and stationary samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Extraction of SSNMR parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chlorine-35/37 experimental data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Simple chlorides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Chlorine in non-cubic inorganic and organometallic compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Organic chlorides and hydrochlorides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Perchlorates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Glasses and sodalites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bromine-79/81 experimental data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Simple cubic bromides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Perbromates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Other materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

216 217 217 217 217 218 219 220 220 221 222 222 222 225 227 228 229 229 231 231

Abbreviations: Cp, cyclopentadienyl; Cp*, pentamethylcyclopentadienyl; CP/MAS, cross-polarization magic-angle spinning; CQ, quadrupolar coupling constant; CS, chemical shift; CSA, chemical shift anisotropy; CT, central transition; d11, d22, d33, principal components of the chemical shift tensor; EFG, electric field gradient; gQ, quadrupolar asymmetry parameter; HMS, Haeberlen–Mehring–Spiess; IUPAC, International union of pure and applied chemistry; j, skew of the magnetic shielding (or chemical shift) tensor; MAS, magic-angle spinning; MQMAS, multiple-quantum magic-angle spinning; NQR, nuclear quadrupole resonance; PAS, principal axis system; Q, nuclear electric quadrupole moment; QCPMG, quadrupolar Carr–Purcell–Meiboom–Gill. * Corresponding author. Tel.: +1 613 562 5800x2018; fax: +1 613 562 5170. E-mail address: [email protected] (D.L. Bryce). 0079-6565/$ - see front matter Crown Copyright Ó 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pnmrs.2009.05.001

216

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

6. Iodine-127 experimental data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Simple cubic iodides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Periodates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3. Other materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Outlook and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Chlorine, bromine, and iodine are elements of importance in a variety of inorganic and organic compounds. For example, chlorine plays an important biological role in vital chloride ion channels [1– 4] and all three elements are present in a variety of inorganic catalysts [5,6]. Solid-state nuclear magnetic resonance (SSNMR) spectroscopy of the chlorine-35/37, bromine-79/81, and iodine-127 nuclides is often challenging due to their NMR properties; however, the experiment is an excellent probe of the local halogen environment in many solid compounds and materials because of the technique’s sensitivity to the local electronic and structural environments. Chlorine-35/37 (I = 3/2), bromine-79/81 (I = 3/2), and iodine127 (I = 5/2) are all NMR-active quadrupolar nuclei that are present in relatively high natural abundances (Table 1). As these nuclei are

231 233 233 234 234 235 235

quadrupoles, the collection of high signal-to-noise (S/N) ratio SSNMR spectra may be difficult, especially for powder samples. In systems where the nucleus does not sit at a site of very high symmetry (e.g., octahedral or tetrahedral), the interaction between the quadrupole moment (Q) of the nucleus and the electric field gradient (EFG) tensor is a significant perturbing interaction to the Zeeman Hamiltonian, which broadens the resulting NMR spectrum of a powder. As halogen atoms do not generally sit at a site of high symmetry in the majority of systems of interest, quadrupolar broadening is generally the dominant contribution to the observed NMR line widths. In these cases, often only the central transition (CT, m = ½ M ½) is observed; even this may be broadened to a significant extent, up to the order of magnitude of MHz in some cases. Further discussion of the NMR properties of the quadrupolar halogen nuclei is given in Section 2.3. Despite the challenges, there have been a number of significant studies using chlorine, bromine,

Table 1 NMR properties of the quadrupolar halogens (chlorine, bromine, and iodine).

35

Cl Cl 79 Br 81 Br 127 I 37

a b c

Natural abundance/%

I

c/107 rad T1 s1

Na

Q/mb

1c1[17]

Relative CT line widthb

IUPAC chemical shift referencec

75.78 24.22 50.69 49.31 100.0

3/2 3/2 3/2 3/2 5/2

2.624198 2.184368 6.725616 7.249776 5.389573

9.797909 8.155725 25.05390 27.00658 20.00748

81.65(80) 64.35(64) 313(3) 262(3) 696(12)

42.0 42.0 80 80 162

1.34 1.00 7.70 5.01 11.44

0.1 mol/dm3 NaCl in D2O 0.01 mol/dm3 NaBr in D2O 0.01 mol/dm3 KI in D2O

Resonance frequency, in MHz, in a magnetic field where 1H resonates at exactly 100 MHz. For a powder sample, based on Eq. (13). From reference [177].

Fig. 1. Approximate chemical shift ranges for chlorine, bromine, and iodine. Scales are with respect to solid NaX (X = Cl, Br, and I) at 0 ppm. Note the different scale for each element.

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

and iodine SSNMR, and with the availability of high (and ultrahigh, P18.8 T) magnetic fields, the amount of research occurring in this area is expected to grow. Presented in Fig. 1 are the approximate solid-state chemical shift ranges for 35/37Cl, 79/81Br, and 127I in various types of compounds. It is apparent that the majority of studies have been performed on halides and perhalogenates, due to the high symmetry (and corresponding small or negligible EFG) at the halogen nuclei in these compounds. This article is a concise overview of chlorine, bromine, and iodine SSNMR, discussing the experimental data from selected studies of diamagnetic samples, with emphasis on the most recent advancements in the field. Two reviews of chlorine-35/37, bromine-79/81, and iodine-127 SSNMR have previously been published. The first surveys the literature up to July 2005 [7], while the second provides a comprehensive examination of this field, up to August 2008 [8]. There have also been reviews of solution-state chlorine, bromine, and iodine NMR by Lindman and Forsén in 1976 [9] and 1978 [10], Forsén and Drakenberg in 1983 [11] and Akitt in 1987 [12]. We emphasize that quadrupolar interaction (QI) data for 79/81Br and 127I in particular are available for a wide range of compounds, primarily via NQR spectroscopy and Zeeman-perturbed NQR experiments. We refer the reader to Lucken’s monograph [13] and to our most recent prior review [8] for more comprehensive accounts of the data obtained from these techniques.

In Eq. (2), m and mref represent the resonance frequency of the sample and the reference (in the same B0), respectively. Both the magnetic shielding and CS are second-rank tensor quantities which, when diagonalized, have three principal components (i.e., eigenvalues) associated with their principal axis system (PAS; i.e., eigenvectors). Thus, Eq. (1) may be used to calculate each element of a chemical shift tensor (dij), from the corresponding magnetic shielding tensor element (rij) of the sample, if the isotropic magnetic shielding value of the reference is also known. The principal components of the chemical shift tensor are defined such that d11 P d22 P d33. The convention for reporting chemical shift anisotropy (CSA) data is not uniform within the literature. Popular conventions include the Haeberlen–Mehring–Spiess [20–23] convention and the Herzfeld–Berger/Maryland convention [21,24]. A brief summary of both conventions is therefore warranted. Where we have summarized CS tensor data in this article, we have reported the principal components, when possible. 2.1.1. Haeberlen–Mehring–Spiess (HMS) convention There are three parameters in the HMS convention that are defined in terms of the three shielding tensor principal components: rXX, rYY, and rZZ. These parameters are the isotropic shielding value (riso), the shielding tensor anisotropy (Dr), and the shielding tensor asymmetry (g), which are defined below:

2. Theory, conventions, and chemical shift referencing

riso ¼ ˆ ) for a NMR-active nucleus within a The Hamiltonian operator (H solid sample contains a term arising from the nuclear magnetic ˆ r) interaction, which acts as a perturbation to the Zeeshielding (H ˆ Z). The latter term arises from the interaction of the man term (H nuclear magnetic moment with the external applied magnetic field, B0. In the case of the quadrupolar chlorine, bromine, and iodine nuclides, the Hamiltonian contains a quadrupolar term as well ˆ Q). There are other terms, e.g., direct homonuclear and heteronu(H clear dipolar coupling, as well as indirect (J) spin–spin coupling, which may be present, but these have a negligible impact on the chlorine, bromine, and iodine SSNMR spectra for most of the compounds studied to date. Thus, only the nuclear magnetic shielding and QI will be briefly described below for the case of diamagnetic powders. More complete theoretical treatments are available and for these the reader is referred elsewhere [13–19]. 2.1. Magnetic shielding interaction and chemical shifts In powdered solids, the averaging of the magnetic shielding tensor (r) that occurs in liquids, due to rapid, isotropic molecular tumbling, is absent and therefore the shielding is not averaged to its isotropic value (riso). In the absence of magic-angle spinning (MAS), the shielding arising from all molecular orientations in a powdered sample contribute to the SSNMR spectrum to produce an anisotropic ‘powder pattern’. While it is not generally possible to measure shielding constants directly in conventional NMR experiments, it is routine to measure the shielding difference between two compounds, with one serving as a reference. This ‘‘chemical” shift (CS, d) can be defined in terms of magnetic shielding as:

d¼

rref  r ; 1  rref

m  mref : mref

rXX þ rYY þ rZZ 3

ðrXX þ rYY Þ 2 3ðrYY  rXX Þ g¼ 2 Dr

Dr ¼ rZZ 

ð3Þ ð4Þ ð5Þ

where jrZZ  riso j P jrXX  riso j P jrYY  riso j. The sign of Dr therefore depends on the size of rZZ relative to riso, and is positive if rZZ > riso and negative if rZZ < riso. This fact can complicate the comparison of Dr values across a range of compounds. The asymmetry is always positive, taking a value between 0 and 1. 2.1.2. Herzfeld–Berger/Maryland convention The three principal components of the shielding tensor, r11, r22, and r33, are always arranged as follows: r11 6 r22 6 r33. As in the HMS convention, the isotropic shielding value is the average of the three principal components: (1/3)(r11 + r22 + r33). The other parameters used to describe the magnetic shielding tensor are the span (X), which describes the breadth of the pattern, and the skew (j), which describes the asymmetry. These parameters are defined below:

X ¼ r33  r11 ;

j¼

ð6Þ

3ðriso  r22 Þ

X

:

ð7Þ

The span can take on any positive value, while the skew ranges between 1 and +1. One advantage of the Herzfeld–Berger/Maryland convention is the direct comparison between the CS and the magnetic shielding tensor. The definitions of the three parameters can be applied to the CS tensor, as shown below:

ð1Þ

where r and rref represent the magnetic shielding of the sample and the reference, respectively. The CS may be more practically defined in terms of resonance frequencies in lieu of shielding constants:

d¼

217

ð2Þ

d11 þ d22 þ d33 ; 3 X  d11  d33 ; diso ¼

j¼

3ðd22  diso Þ

X

:

ð8Þ ð9Þ ð10Þ

As long as 1–rref is close to unity, Eq. (9) is valid, allowing for the direct comparison of chemical shift and (often calculated) magnetic

218

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

shielding data. This is practically always the case when discussing chlorine, bromine, and iodine SSNMR data. 2.2. Nuclear electric quadrupolar interaction As chlorine, bromine, and iodine are all quadrupolar nuclei, the nuclear charge is distributed asymmetrically about the nucleus. The resultant Q will couple with the EFG tensor (V) at the nuclear centre. The EFG tensor is traceless and symmetric and may be described by five independent parameters. In its PAS, the three principal components of the tensor are labelled, V11, V22, and V33, with magnitudes arranged as shown:

jV 33 j  jV 22 j  jV 11 j: Three Euler angles describe the orientation of the EFG PAS with respect to an external axis system. Since the EFG tensor is traceless, the principal components may be fully described using two independent parameters. These components are generally expressed in terms of the quadrupolar coupling constant (CQ) and the asymmetry parameter (gQ):

eV 33 Q ; h V 11  V 22 gQ ¼ ; V 33 CQ ¼

ð11Þ ð12Þ

where h is Planck’s constant and e is the elementary charge. The quadrupole moment is a constant for a given nucleus (Table 1). To fully describe the CS and EFG tensors, the orientations of their PASs must be known. In practice, for powder samples one typically measures the relative orientation of the PASs of the CS and EFG tensors. This requires an additional three parameters, known as the Euler angles, a, b, and c. Under the Arfken convention [25], these angles describe the counter-clockwise rotations of the CS tensor away from the stationary EFG reference tensor, assuming initial coincidence of the two PASs. Specifically, a describes the initial rotation about the original d33 axis, b is a rotation about the new direction of d22, and c is the final rotation, around the new direction of d33, as shown in Fig. 2. Chlorine, bromine, and iodine all have moderate to high Q values and thus the quadrupolar interaction is often quite significant. Hence, first-order perturbation theory generally fails to describe the SSNMR line shapes which are observed experimentally for powdered samples. However, second-order perturbation theory

usually adequately describes the experimental SSNMR line shape. Generally in the case of these nuclei, the QI is sufficiently large to broaden the satellite transitions (ST) to such an extent that they cannot be observed, and experimentally only the CT can be detected. The CT is not broadened by first order quadrupolar effects but is affected in the second order. The breadth of the CT line shape of a stationary powdered sample depends on four parameters (CQ, m0, I, and gQ), as shown below [17]:

DmCT ¼

3C Q 2Ið2I  1Þ

2

ðg2Q þ 22gQ þ 25ÞðIðI þ 1Þ  34Þ ; 144m0

ð13Þ

where m0 is the Larmor frequency. The inverse proportionality between DmCT and m0 demonstrates why higher magnetic fields are advantageous for the quadrupolar halogens: increasing B0 increases m0, and therefore results in a decrease in the second-order quadrupolar broadening of the CT. Presented in Table 1 are the relative values of DmCT for the five quadrupolar halogen nuclides for a given EFG; these are calculated directly from Eq. (13). It is also worth noting that the effects of CSA increase at higher fields, allowing for these effects to be more easily observed in cases where there is a significant QI. In addition, for a given EFG tensor, narrower CT powder patterns are obtained for higher half-integer spin nuclei (see Eq. (13)). The Sternheimer antishielding factor, 1–c1, modifies the EFG ‘felt’ at the nucleus due to the polarization of inner-shell electrons [13,26,27]. That is, the EFG due solely to ionic charges in the lattice, eqionic, is modified as follows [28]:

eqobserved ¼ ð1  c1 Þeqionic :

ð14Þ

The quantity eqobserved must be considered for the accurate discussion of quadrupolar coupling constants and relative line widths (vide supra). The antishielding factors for Cl, Br, and I are tabulated in Table 1. Wu and Terskikh have recently demonstrated the importance of considering the antishielding factors when comparing NMR data for alkali metal cations in isostructural tetraphenylborate salts [28]. The importance for chlorine, bromine, and iodine may also be seen by considering that 1–c1 increases from 42.0 to 80 to 162 for these three elements. We may appreciate the impact of antishielding by considering the following: assuming each of the nuclides in turn is hypothetically placed into the same crystal lattice of fixed geometry and volume, they will experience identical electric field gradients due to the charges in the lattice (eqionic). However, the EFG at the nucleus will differ due to the different values of the antishielding factors, as per Eq. (14). The relative impact due to antishielding on the value of DmCT may be assessed by evaluating [Q(1–c1)]2/c. The results for an axially symmetric EFG are 44.8, 33.4, 930, 610, and 23600  1064 m4 rad1 T s for 35Cl, 37Cl, 79 Br, 81Br, and 127I, respectively. This highlights the challenge of observing 127I NMR spectra even in chemical environments where chlorine and bromine NMR spectra have been successfully acquired. Of course, the above assessment does not take into account the nuclear spin quantum numbers, which also affect the line widths, according to Eq. (13). For the majority of the results discussed in this work, the highfield approximation, which assumes that m0 is much greater than the quadrupolar frequency (mQ),

mQ Fig. 2. Depiction of the Euler angles a, b, and c used for relating the principal axis systems (PASs) of the electric field gradient tensor (V) and the chemical shift tensor (d). The angles may also be visualized as counterclockwise rotations of the d tensor PAS away from the V PAS: first about the eigenvector corresponding to d33 through an angle a, then about the intermediate direction of the d22 eigenvector by an angle b, then about the final direction of the d33 eigenvector by an angle c.



qffiffiffiffiffiffiffiffiffiffiffiffiffi2 g 3C Q 1 þ 3Q ¼ 2Ið2I  1Þ

ð15Þ

is used. The validity of this approximation in the case of nuclei with large quadrupole moments depends on the EFG tensor at the nuclear site in a given compound; the value of mQ may become comparable to the Larmor frequency in certain cases. The range of validity was recently tested through several simulations that highlighted the differences between the second-order and exact approaches

219

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

[8,29]. It was found that the difference between the calculated exact and second-order perturbation theory CT resonance frequencies for a single crystal is negligible (<1%) for practical purposes up until the point where mQ is m0/2 for I = 3/2, and m0/5 for I = 5/2. The difference is even smaller in the case of the CT of a powdered sample. These results suggest that the common rule-of-thumb regarding the validity of the second-order approximation (i.e., mQ < 0.1m0) may be slightly conservative, particularly for powdered samples [8]. 2.3. NMR properties of chlorine, bromine, and iodine Of the quadrupolar halogens, chlorine is the most ‘‘NMR friendly”, with its two NMR-active isotopes having quadrupole moments significantly lower than those of both bromine and iodine (Table 1). Although the quadrupole moment of 37Cl is lower, 35Cl is the more commonly studied isotope due to its higher natural abundance. The moderately low Larmor frequencies of 35Cl and 37 Cl can present some problems with acoustic ringing in low magnetic fields, but with the increased availability of higher field instruments, chlorine experiments are becoming more accessible. 35 Cl resonates just below 15N and so it is often possible to observe 35 Cl on a standard HCN probe. Compared to chlorine, 79/81Br and 127 I have higher Larmor frequencies and their NMR-active isotopes exist in high natural abundances; however, their moderately large quadrupole moments and the concomitant spectral broadening make collecting spectra of these nuclei difficult, except in cases of very high symmetry. Of the three halogens, 127I is generally the most difficult nucleus to study experimentally, despite its 100% natural abundance, higher value of I, and moderate m0, due to its large Q value and the typical electronic environments in which 127I is found. Solution-state NMR standards for each of the three nuclei have been recommended [30], as shown in Table 1. When using these standards, the concentration must be carefully determined and noted as the chemical shift may vary with concentration [31]. The solvent (e.g., H2O vs D2O) is even more important, as significant isotope effects are known to exist [7,32]. For example, for a 1–3 M NaCl solution, d(35Cl, H2O)–d(35Cl, D2O) = 4.7 ± 0.3 ppm; for a 0.41– 0.82 M RbBr solution, d(79/81Br, H2O)–d(79/81Br, D2O) = 8.3 ± 0.3 ppm and for a 0.45–0.95 M NaI solution, d(127I, H2O)–d(127I, D2O) = 13.2 ± 2.5 ppm [33]. The alternative references of solid NaCl or KCl have been suggested for 35/37Cl, and solid KBr for 79/81Br, as they produce narrow lines at relatively slow spinning speeds because of their cubic structures. In addition, this removes any ambiguity in terms of concentration or solvent isotope effects [7,8]. To facilitate chemical shift referencing using solid powdered alkali metal halides, we have measured their chemical shifts with respect to the IUPAC standards (Table 2). Shown in Figs. 3 and 4 are chlorine-35/37 and iodine-127 SSNMR spectra of solid powdered alkali metal halides with respect to the solution-state IUPAC stan-

Fig. 3. Chlorine-35 MAS NMR spectra of powdered alkali metal chlorides (B0 = 9.4 T). Spectra of the individual samples are shown in the top traces, with the centrebands labelled. Spectrum of a mixture of the five compounds is also shown (bottom). The chemical shifts are referenced with respect to the IUPAC standard, 0.1 mol dm3 NaCl in D2O at 0 ppm. (see Table 2 for chemical shifts).

dards. The 79Br and 81Br chemical shifts of the alkali metal bromide salts, relative to the IUPAC solution standard, were reported by Widdifield et al. [8] All the data are presented in Table 2. Of the nuclei under discussion, chlorine is the only one for which a precise absolute shielding scale is available. This scale was determined through a combined experimental and theoretical study. An isotropic magnetic shielding constant (riso) of 974(4) ppm was determined for a chloride ion in an infinitely dilute aqueous solution [34]. The conversion between shielding and shift is therefore as follows (cf. Eq.(1)):

d ¼ ð974ppm  rÞ=ð1  974ppmÞ

ð16Þ

and is an improvement over prior chlorine absolute shielding scales [35,36]. The bromine and iodine absolute shielding scales are much less precisely determined and in fact no references have been made to

Table 2 Solid-state halogen isotropic chemical shifts of the alkali metal halides, MX.a With respect to KX(s)

LiX NaX KX RbX CsX

With respect to IUPAC aqueous standard

M35Clb

M79Brb

M127Ib

M35Clc

M79Brd

M81Brd

M127Ie

1.31 49.73 0 41.13 105.96

64.74 52.89 0 71.66 227.43

215.28 33.53 0 76.91 368.69

9.93 41.11 8.54 49.66 114.68

119.06 1.29 54.31 125.89 282.66

119.33 1.57 54.51 126.13 282.76

408.66 226.71 192.62 269.87 562.41

a Slight differences (<1 ppm) between relative shifts with respect to the solid and the solution are seen; this is likely due to the fact that the data come from two different sources. b With respect to KX(s). From references [69,74]. c With respect to 0.1 M NaCl in D2O. d With respect to 0.03 M NaBr in D2O. From reference [8]. e With respect to 0.01 M KI in D2O.

220

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Fig. 4. Iodine-127 MAS NMR spectra of powdered alkali metal iodides (B0 = 9.4 T). Spectra of the individual samples are shown in the top traces, with the centreband labelled. A spectrum of a mixture of the five compounds is also shown (bottom). The chemical shifts are referenced with respect to the IUPAC standard, 0.01 mol dm3 KI in D2O at 0 ppm (see Table 2 for chemical shifts).

the bare nuclei. Hartree–Fock methods using a semi-quantitative theoretical model [37] in combination with experimental data [38,39] were used to calculate the shift of the bromide anion in dilute aqueous solution with respect to the free gas-phase anion as 194 ppm [40]. For I in dilute aqueous solutions, a value of 600 ppm with respect to the free anion was determined by Itoh and Yamagata via experimental solution-state measurements [41]. It is non-trivial to establish a reliable absolute shielding scale for bromine and iodine, as it is not clear how the non-relativistic relationship between the spin-rotation tensor and the magnetic shielding tensor (which is a common method of establishing an absolute shielding scale for a particular nuclide) is affected when relativistic effects become important [42,43]. 3. Experimental methods for powders 3.1. Magic-angle spinning and stationary samples The majority of SSNMR studies on chlorine, bromine, and iodine are done on powdered samples. With powdered samples, MAS experiments are a powerful method of study in cases when the broadening of the CT is small enough to allow for the acquisition of a MAS NMR spectrum with resolution of spinning sidebands (ssb) from the centreband. In these cases, the resultant centreband line shape is free of any CSA effects. The second-order QI is not completely averaged by the spinning, thereby allowing for the extraction of CQ and gQ, along with diso, as these are the three parameters that quantify the line shape and position of the centreband. As the broadening associated with the QI decreases with increasing field (Eq. (13)), the increased availability of high/ultrahigh–field magnets has allowed for the acquisition of MAS spectra

for materials with a greater range of CQ values. In addition, advances in probe technology allow for higher spinning rates and have also opened the door to studying a variety of materials with more strongly quadrupolar-broadened NMR lineshapes. Currently, the highest magnetic field readily available for SSNMR studies is 21.84 T and the highest commercially available MAS frequency is 70 kHz, using a 1.3 mm o.d. rotor. Two-dimensional (2D) NMR experiments have been developed to collect high-resolution ‘‘isotropic” spectra of half-integer quadrupolar nuclei in which the quadrupolar effects have been removed from the MAS spectrum. These include the multiple-quantum MAS (MQMAS) and satellite transition MAS (STMAS) pulse sequences. Both 2D sequences remove the quadrupolar broadening in one dimension through excitation of either multiple-quantum coherences (in MQMAS)[44] or satellite transitions (in STMAS) [45], followed by a transfer of the magnetization to the CT for detection and appropriate data processing. To the best of our knowledge, the STMAS experiment has not been applied to chlorine, bromine, or iodine. There exists one example of MQMAS applied to the quadrupolar halogens: Trill et al. report the collection of a 35Cl MQMAS spectrum in their study of chlorine-containing sodalites, although the spectrum is not shown in their publication [46]. While both pulse sequences are very useful for a variety of nuclei, it appears that their application to the quadrupolar halogens is hindered due to the large QI in many of the materials of interest. A common experimental technique used to improve the quality of the CT spectrum for the quadrupolar halogens is a simple quadrupolar echo. Due to the large Q values for the quadrupolar halogens, the line shapes in the resulting spectra are often very broad and therefore have rapidly decaying free induction decays (FIDs). With chlorine, this may be coupled with the presence of acoustic ringing in the coil, which can prevent the measurement of the resulting signal. This problem is remedied by a quadrupolar echo, as it shifts the FID forwards in the time domain. The quadrupolar Carr–Purcell–Meiboom–Gill (QCPMG) pulse sequence [47] is also used to improve experimental sensitivity and has started to be applied to chlorine and bromine [48–53]. The experiment consists of the collection of the FID after each pulse in a train of refocusing p pulses after an initial p/2 pulse. After Fourier transformation, the frequency-domain spectrum is a powder pattern which is split into ‘‘spikelets”. The experiment can be carried out under stationary, MAS, and MQMAS conditions and improves experimental sensitivity on two fronts: firstly, the multiple signal acquisition periods per relaxation delay and secondly, the enhancement from concentrating the total signal intensity into the spikelets. The improved sensitivity greatly reduces the acquisition time needed to collect a spectrum with adequate S/N, compared to a spin echo. Siegel et al. used the QCPMG experiment to acquire a 127I NMR spectrum of KIO4, a compound which has a CQ value over 20 MHz (Fig. 5) [54]. We have found this compound to be an excellent set-up sample for 127I QCPMG NMR experiments. Other signal enhancement methods are available to increase the S/N ratio of the spectrum of the CT. A popular approach is to ‘borrow’ intensity from the satellite transitions either through saturation or inversion of their populations [55]. Implementations of these methods, such as double-frequency sweeps (DFS)[56] and hyperbolic secant pulses [57], have begun to be applied to the quadrupolar halogens, and they will likely prove useful in future experiments. For example, Bryce et al. have reported applications of the hyperbolic secant enhancement method for 35Cl NMR studies of amino acid hydrochlorides [49]. In some instances, the line width of the central transition is so large that it is not possible for it to be uniformly excited. In this case, variable-offset experiments are useful to collect the full SSNMR spectrum. The approach of these experiments is simple: the rf transmitter is systematically stepped across the entire

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

221

Fig. 5. Iodine-127 NMR spectra of solid powdered KIO4 obtained under stationary conditions at 11.75 T. Pulse sequence used: (a) QCPMG; (b) modified QCPMG (the repeating p pulses are replaced with p/2 pulses); (c) converging DFS with QCPMG; (d) converging DFS with modified QCPMG. The relative S/N ratios are given on the right. From R.S. Siegel, T.T. Nakashima, R.E. Wasylishen, Concepts Magn. Reson. Part A 26A (2005), 62. Reproduced by permission of Wiley Inc.

breadth of the CT, with a ‘subspectrum’ being acquired at each offset. The series of subspectra are then overlapped (skyline processed) or summed (co-addition processed) to produce the entire CT spectrum. In cases where the QCPMG method is used, the transmitter frequency offsets must be chosen so that the spikelets in different experiments overlap constructively. In our experience, the QCPMG sequence has been advantageous to achieve the needed S/N ratio for variable-offset experiments involving 35Cl, due to its lower receptivity, while simple echoes are feasible for the bromine nuclides when they are present in ionic systems [51]. An example of the skyline processing method is presented in Fig. 6, where the

QCPMG variable-offset approach has been applied to acquire the Cl CT spectrum of indium trichloride tetrahydrate at 21.1 T. This method greatly expands the range of chlorine, bromine and iodine materials that can be studied with SSNMR, as experiments are not limited by the excitation bandwidth of the probe. In the future, the WURST-QCPMG method of O’Dell and Schurko, which allows for the efficient acquisition of broad spectra, will also be useful for the collection of the broad SSNMR spectra of 35/37Cl, 79/81Br, and 127 I [58,59]. Other methods have been used in select chlorine, bromine, or iodine SSNMR studies. Satellite-transition spectroscopy (SATRAS), where the MAS NMR spectrum of the satellite transitions is collected and analyzed [60–62], was used by Trill et al. on a series of halide-containing sodalites [46]. More recently, Persons and Harbison collected 79Br SSNMR spectra of several samples using their ‘‘slow turning reveals enormous anisotropic quadrupolar interactions” (STREAQI) method [63]. The STREAQI method was shown to characterize the quadrupolar parameters of sites with moderately large QIs (10–50 MHz) using very slow MAS rates (e.g., 300 Hz), although the current analysis method does not take into account CSA. 35

3.2. Extraction of SSNMR parameters

Fig. 6. Variable-offset chlorine-35 CT NMR spectrum of InCl3  4H2O. The 25 QCPMG subspectra have been overlaid to give the shape of the overall spectrum, which covers about 1 MHz at 21.1 T. The solid line is an analytical simulation from which the following approximate parameters may be extracted: CQ(35Cl) = 23.4 MHz; gQ = 0.27; diso = 130 ppm. Decreasing the offset used would reduce the oscillations in the envelope of the experimental pattern.

As mentioned above, the centreband of the CT under fast MAS conditions depends on diso, CQ, and gQ and these parameters can typically be determined by line shape analysis (as implemented in WSolids [64,65], SIMPSON [66], or DMFIT [67], for example). Thus, where possible, the desired method to extract both the CS and QI tensor parameters for a sample is to first analyze the MAS spectrum to extract diso, CQ, and gQ and then run a static experiment and perform a line shape analysis of that spectrum to extract the remaining parameters: X, j, a, b, and c. In cases where the QI is too great to obtain a MAS spectrum, the static spectrum must generally be fit with all eight parameters as variables. As there is often more than one combination of the eight parameters that can give

222

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

the same spectrum, collection of the spectrum at multiple fields is preferred, in order to confirm the accuracy of the determined parameters, as the same set of parameters must fit all the spectra acquired for a given material. Conveniently, in the case of chlorine and bromine there is a second ‘‘built-in” field, as there are two NMR-active isotopes, both in reasonable natural abundances, of those two elements. To a good approximation, the only parameter that changes when comparing the two isotopes of chlorine or bromine, is the CQ value, due to each possessing a different Q value. In the case of chlorine, the chlorine-37 coupling constant is 78.8% that of the chlorine-37 value, while the bromine-81 coupling constant is 83.7% that of the value for bromine-79 [68]. In principle, one must be aware of possible isotope effects on the chemical shift, but these are expected to be negligible, particularly in the analysis of broad powder patterns. 4. Chlorine-35/37 experimental data The types of materials that have been analyzed using 35/37Cl SSNMR include biological molecules, inorganic materials, and chlorine-containing glasses [7]. The chlorine chemical shift range is approximately 1100 ppm, ranging from 78.06 for CuCl [69] to 1049.3 ppm for KClO4 [70] (both referenced to NaCl(s)) (Fig. 1). Of the three quadrupolar halogens, the effects of CSA are most visible in chlorine NMR spectra as a result of the smaller Q value for both isotopes of chlorine. Thus, only for the chlorine nucleus have there been a significant number of samples for which the CS tensor has been characterized [7]. Chlorine CSA data are summarized in Table 3. 4.1. Simple chlorides Selected SSNMR data for inorganic chloride salts are given in Tables 2 and 4. These tables, along with all other data tables in this review, are not intended to be comprehensive. They represent an overview of the available single crystal and powdered sample data, emphasizing studies we feel are significant to the field. In the case of the simple chlorides, this entails including those studies that provided accurate chemical shift data referenced to relatively common reference compounds. More comprehensive data tables and detailed discussion pertaining to several metal halides are available in our prior review [8]. Simple chloride salts have been popular materials of study by 35/37 Cl SSNMR, due to the cubic environment about the chlorine nucleus in many such compounds. The high symmetry of these systems results in small CQ(35/37Cl) values, which allows them to be analyzed in relatively low magnetic fields. The first chlorine SSNMR study appears to have been done by Kanda in 1955 on powdered samples of four simple salts: NaCl, CsCl, AgCl and TlCl [71]. The relative order of the chemical shifts was reported and rationalized by the suggestion that the shift increased with increasing bond covalency. In addition, the study demonstrates the ability to study these cubic materials at low magnetic fields (i.e., 0.300– 0.622 T). A comprehensive study of all the alkali metal chlorides was performed using powdered samples by Yamagata in 1964 and a chemical shift trend was noted [72]. This study found that, with the exception of lithium, the chlorine CS increased upon increasing the atomic number of the alkali metal. Yamagata concluded from this study that the dominant factor in determining the chlorine chemical shift was the orbital overlap between the halogen ion and its nearest and next nearest neighbours. Additional comprehensive studies of cubic chloride salts were carried out several decades later. Weeding and Veeman studied NaCl, KCl, RbCl, CsCl and NH4Cl [73]. The investigators collected MAS NMR spectra of powdered samples, and observed sharp line-

shapes in which second-order quadrupolar broadening was absent. The absence of quadrupolar broadening demonstrated the absence of a QI, as expected due to the cubic arrangement of the ions in these salts. Both isotopes of chlorine were analyzed by the authors, and were found to have nearly identical chemical shifts, as would be expected in the absence of significant isotope effects. The chlorine chemical shift was found to correlate with the M–Cl bond distance, as well as the Sanderson electronegativity of the cation. Shortly thereafter, Hayashi and Hayamizu [74] acquired both MAS and static chlorine NMR spectra for seven simple chloride salts: the alkali metal chlorides (LiCl, NaCl, KCl, RbCl, and CsCl), as well as CuCl and AgCl. The authors rationalized the chlorine chemical shifts in a similar manner to Yamagata, focusing on the orbital overlap between the nucleus and its neighbours [72]. In addition to the chlorine NMR spectra, the authors collected the 7 Li, 23Na, 87Rb, and 133Cs MAS NMR spectra of the appropriate salts. In 1992, Lefebvre reported 35Cl chemical shifts for selected simple salts under static conditions [75]. The relationship between chemical shift and Sanderson electronegativity, noted previously [73], was confirmed. The possibility that earlier researchers were actually collecting the 35Cl NMR spectrum for a hydrate of lithium chloride, Li(H2O)x, was proposed as an explanation for the anomalous chlorine chemical shift observed for ‘LiCl’. Honda examined several n-alkylammonium chlorides, nCxH(2x+1)NH3Cl [76]. Some salts were deuterated, and it was found that in all cases, these salts had lower magnitudes of CQ(35Cl), compared to their non-deuterated counterparts. It was also noted that the chlorine-35 CQ followed an ‘even-odd’ trend in the highest temperature phase of the salts. The ‘even’ salts (x = 8 and 10) had much larger coupling constants compared to those of the ‘odd’ salts (x = 5, 7, 9). All CQ(35Cl) magnitudes fell in the range of 1.0– 1.5 MHz. Chlorine-35 NMR spectra of these salts and their deuterated counterparts are shown in Fig. 7. There have been numerous additional studies on both single crystals and powdered samples of the simple chloride salts, and a variety of properties were measured [8]. Examples include using the temperature dependence of T1 in NaCl and AgCl to calculate defect migration, the pressure and temperature dependence of diso(35Cl) in RbCl, and the effect of elastic strain on the chlorine CT line shape in NaCl, amongst others [39,77–81]. 4.2. Chlorine in non-cubic inorganic and organometallic compounds Chlorine quadrupolar parameters for non-cubic inorganic and organometallic compounds are presented in Table 4 and available CS tensor parameters are summarized in Table 3. Rossini et al. [52] very recently presented an extensive 35Cl SSNMR study of a series of group IV transition metal organometallic complexes, including Cp2TiCl2, CpTiCl3, Cp2ZrCl2, Cp2HfCl2, Cp*2ZrCl2, CpZrCl3, Cp*ZrCl3, Cp2ZrMeCl, and (Cp2ZrCl)2(l-O) (Cp = cyclopentadienyl; Cp* = pentamethylcyclopentadienyl). Signal-enhancement techniques such as QCPMG, combined with variable-offset data acquisition and ultrahigh magnetic fields (21.1 T), were applied in this study. The values of CQ(35Cl) were found to range from 12.8(5) MHz for one of the four chlorine sites in Cp*ZrCl3 to 22.1(5) MHz in Cp2TiCl2. Although this study is apparently the first to characterize chlorine in a ‘‘covalent” environment using powdered samples and modern high-field NMR methods, we note that the CQ values are significantly lower than those known for covalently-bound chlorine atoms in organic molecules. For example, 35Cl NQR data for a range of chlorine-containing organic heterocycles (e.g., 2-chloropyridine) indicate quadrupolar coupling constants on the order of 70 MHz [13]. An indirect NMR measurement of CQ(35Cl) in 2-chloro-2(phenylsulfonyl)-1-phenylpropanone, obtained by measuring the residual dipolar coupling in the 13C CP/MAS NMR spectrum, yielded a value of 73 ± 2 MHz [82]. In this context, the values of

223

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237 Table 3 Chlorine-35 chemical shift tensor data.a Compound

diso/ppm

d11/ppm

CpTiCl3 Cp2ZrCl2 Cp*ZrCl3 (sites 1 & 2) Cp*ZrCl3 (sites 3 & 4) Butyldimethylimidazolium chloride Butylmethylimidazolium chloride Procaine HCl Tetracaine HCl Lidocaine HCl Lidocaine HCl polymorph 1,f site 1 Lidocaine HCl polymorph 1,f site 2 Bupivacaine HCl Bupivacaine HCl heated to 120 °C, site 1 Bupivacaine HCl heated to 120 °C, site 2 Bupivacaine HCl heated to 170 °C Alanine HCl Aspartic acid HCl Cysteine HCl monohydrate Histidine HCl monohydrate Methionine HCl Threonine HCl MgCl2  6H2O CaCl2  2H2O SrCl2 SrCl2  2H2O SrCl2  6H2O BaCl2  2H2O site 1 BaCl2  2H2O site 2

500(150)c 300(150)c 400(200)c 200(200)c 71.60(0.02)d 71.65(0.05)d 96(6)e 71(6)e 100(4)e 85(10)e 110(10)e 96(10)e 118(10)e 95(10)e 118(5)e 106(5)e 102(5)e 104.2(0.5)e 93(1)e 99(1)e 99(10)e 75.0(1.0)e 110.0(2.0)e 188.2(1.0)e 142.1(1.0)e 90.4(1.0)e 163.4(2.0)e 156.6(2.0)e 105.0(1.0)e

925 700 617

L-tryptophan

HCl

101.90 96.66 166.8 105.6 170.6 97.7 126.5 142.7 174 180.3 162 139 150.8 135.9 144 149.7 138.8 188.2 159.2 120.4 195.1 179.9 139.8

d22/ ppm

d33/ppm

400 175 300 100 467 117 X = 200(200) 88.69 24.20 65.13 53.16 79.3 41.8 81.7 25.7 68.8 60.6 79.7 77.7 122.0 81.5 102.7 42.7 166 14 84.3 20.3 150 42 100 79 79.5 75.8 106.8 69.9 X < 150 109 44 92.7 54.7 X < 75 124.4 66.8 188.2 188.2 148.9 118.2 75.4 75.4 150.1 145.1 159.9 129.9 107.4 67.8

a,b,c (deg)b

References

80(30), 5(15), 5(30) 2(10), 72(20), 70(20) 10(90), 15(30), 0(90) – 16(1), 82.5(0.5), 34(4) 78(2), 76(1), 12(2) 95(15), 3(2), 32(8) 60(8), 8(5), 10(10) 12(3), 40(10), 80(3) 90(40), 50(50), 60(40) 5(5), 50(15), 40(40) 105(20), 90(5), 5(5) 10(10), 3(1), 0(2) 18(4), 50(5), 80(5) 10(10), 0(2), 50(50) 90(15), 0(15), 0(15) 0(20), 30(20), 93(20) 155(20), 0(10), 0(20) – 93(20), 163(15), 7(20) 95(15), 0(10), 0(15) – 90(10), 82(5), 0(20) – 86(15), 75(5), 37(10) 0(10), 90(10), 0(10) 85(20), 32(10), 60(20) 20(15), 8(10), 0(20) 90(15), 20(15), 2(20)

Rossini et al. [52] Rossini et al. [52] Rossini et al. [52] Rossini et al. [52] Gordon et al. [90] Gordon et al. [90] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Chapman and Bryce [50] Chapman and Bryce [50] Chapman and Bryce [50] Chapman and Bryce [50] Chapman and Bryce [50] Chapman and Bryce [50] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Sward [48]

91.5(1.0)e

117.7

96.7

60.2

85(15), 77.5(12.0), 30(30)

Bryce and Sward [48]

L-lysine

HCl

105(2)e

119.7

101.5

93.7

0(20), 52(20), 0(20)

Bryce et al. [49]

L-serine

HCl

120(30)e

–

Bryce et al. [49]

DL-arginine

L-glutamic L-proline

HCl monohydrate

acid HCl

HCl

L-isoleucine L-valine

102(1)e

135

102

69

9(20), 77(20), 6(20)

Bryce et al. [49]

37(5)e

74.2

25.7

11.2

48(20), 69(20), 9(20)

Bryce et al. [49]

96(20)e

HCl

HCl

L-phenylalanine

X < 150

HCl

X = 75(30); j < 0.85

20(20), 12(20), 0(20)

Bryce et al. [49]

90(10)e

145.2

104.6

20.2

65(20), 0(20), 0(20)

Bryce et al. [49]

96(5)e

154.9

107.2

25.9

91(20), 13(20), 10(20)

Bryce et al. [49]

146 83.0

111 40.7

46 36.0

95(20), 0(20), 0(20) b = 90

Bryce et al. [49] Bryce et al. [53] Bryce et al. [53]

L-cysteine

ethyl ester HCl

101(5)e 53.2(0.5)c

L-tyrosine

HCl

49.3(0.5)c

X < 150

–

48.2(0.7)c

X = 45(15)

–

Bryce et al. [53]

X = 50(20) X < 150 X ranges from 10 to 160

– –

Bryce et al. [53] Bryce et al. [53] Hou et al. [131,132]; Hou and Kirkpatrick[133]

Glycine HCl

L-cysteine

methyl ester HCl Quinuclidine HCl Tris sarcosine CaCl2 LiAl2(OH)6ClO4  nH2O Mg3Al(OH)6ClO4  nH2O LiAl2(OH)6Cl  nH2O NH4ClO4 p-dichlorobenzene NaClO3 a b c d e f g

9.7(10.0)c 14.7(10.0)c Spectra obtained at several relative humidities 917.5(0.7)c 666.7g –

932.8 918.9 900.8 – 3000(1000) 0(1000) 1000(1000) 0,0,0 (assumed) X = 40(7)

Bastow and Stuart [97] Creel et al. [134] Kawamori and Itoh [135]

Errors are given in parentheses; all parameter definitions are provided in the main text. Euler angles describing the relationship between the PAS of the CS tensor and that of the EFG tensor. With respect to infinitely dilute NaCl(aq). With respect to 0.1 M NaCl in D2O. With respect to NaCl(s). See Ref. [88] for details on sample preparation. No chemical shift reference given.

CQ(35Cl) obtained for the group IV transition metal complexes suggest that the chlorine bonding environment contains more ionic character than in the organic systems. Values of CQ(35Cl) for chlorines in ionic environments within inorganic and organic systems (where Cl is a counterion) can approach 10 MHz (e.g., the CQ(35Cl) value for aspartic acid hydrochloride is 7.1 ± 0.1 MHz [50]). Rossini et al. also provided valuable chlorine chemical shift tensor information for the group IV transition metal chloride complexes [52]. Perhaps most interesting are the isotropic chemical shifts, which fall in between the previously known ranges for chloride ions (approximately 100 to +200 ppm, with respect to

NaCl(s)) and perchlorate ions (approximately 900–1100 ppm). For instance, the value for CpTiCl3 is 545 ± 150 ppm with respect to the same reference. The measured spans of the chlorine CS tensors are quite large relative to other literature values; for example, a value of 800 ± 500 ppm was reported for Cp2ZrCl2. This is consistent with a more covalent bonding environment, relative to the known ionic systems. In the same study, the authors combined 35Cl SSNMR results with X-ray diffraction data and quantum chemical calculations to demonstrate that Cp2ZrHCl (Schwartz’s reagent) exists as a dimer in the solid state [52]. They also attempted to establish definitive

224

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Table 4 Selected chlorine-35 quadrupolar parameters and isotropic chemical shifts for inorganic and organometallic chlorides.a Compound Cp2TiCl2b Cp2ZrCl2 Cp2HfCl2 Cp*2ZrCl2d CpTiCl3 Cp2ZrMeCl (Cp2ZrCl)2(l–O) Cp*ZrCl3 (four sites)

CpZrCl3 (multiple sites) Cp2ZrHCl CaCl2  2H2O BaCl2 BaCl2  2H2O VCl2 MgCl2  6H2O SrCl2 SrCl2 SrCl2  2H2O SrCl2  6H2O Mn(CO)5Cl FeCl2 K2CuCl4  2H2O

BaFCl Ag5Te2Cl a-form b-form c-form a b c d e f

diso/ppm c

500(500) 300(150)c 400(500)c 400(400)c 500(150)c 400(400)c 300(400)c 400(200)c 400(200)c 200(200)c 200(200)c 300c 80(50)c 110.0(2.0)e site 1: 124(5)f site 2: 219(5)f site 1: 163.4(2.0)e site 2: 156.6(2.0)e – 75.0(1.0)e 140.8f 188.2(1.0)e 142.1(1.0)e 90.4(1.0)e – – – –

– 12.9(0.5)f (356 K) 8.9f (296 K) 4.8(0.5)f (216 K)

CQ(35Cl)/MHz

gQ

References

22.1(0.5) 16.0(0.5) 17.1(0.4) 16.7(0.4) 15.5(0.4) 13.7(0.4) 16.3(0.4) 12.8(0.5) 13.3(0.5) 14.6(0.5) 14.0(0.5) 14.8–18.6 19.7(0.3) 4.26(0.03) site 1: 3.5(0.1) site 2: 3.95 site 1: 2.19(0.08) site 2: 3.42(0.08) 4(2 sites) 3.02(0.05) – 0 1.41(0.02) 3.91(0.05) 36.07(0.01) (303 K) 4.74(0.01) 4.74(0.02) (4.2 K) site 1: 18.98 site 2: 3.490 (77 K) site 1: 19.07 site 2: 3.434 (295 K) site 1: 19.05 site 2: 3.262 (340 K) site 2: 3.229 2.38

0.61(0.03) 0.72(0.04) 0.65(0.05) 0.73(0.03) 0.54(0.05) 0.75(0.10) 0.43(0.07) 0.10(0.10) 0.12(0.10) 0.88(0.10) 0.80(0.10) 0.7–0.8 0.20(0.04) 0.75(0.03) 0.15(0.05) 0.1 0 0.31(0.10) 0(assumed) 0 – – 0.80(0.10) 0 0.050(0.005) – – (4.2 K) 0.1883 0.877 (77 K) 0.1902 0.895 (295 K) 0.183 0.945 (340 K) 0.97 0

Rossini Rossini Rossini Rossini Rossini Rossini Rossini Rossini

–

– – –

et et et et et et et et

al. al. al. al. al. al. al. al.

[52] [52] [52] [52] [52] [52] [52] [52]

Rossini et al. [52] Rossini et al. [52] Bryce et al. [83] Stebbins and Du [84] Bryce and Bultz [83] Tabak et al. [136] Bryce and Bultz [83] Lefebvre [75] Bryce and Bultz [83] Bryce and Bultz [83] Bryce and Bultz [83] Spiess and Sheline [137] Barnes and Segel [138] Jones and Segel [139] Choh and Stager [140]

Bastow et al. [141] Brinkmann et al. [142]

Errors are given in parentheses; all parameter definitions are provided in the main text. Cp = cyclopentadienyl. With respect to infinitely dilute NaCl(aq). Cp* = pentamethylcyclopentadienyl. With respect to NaCl(s). With respect to 1 M NaCl(aq).

35 Cl NMR fingerprints for terminal vs. bridging chloride ligands in the systems studied. In the case of Cp*ZrCl3, the bridging chlorine atoms gave rise to a much larger CT powder pattern (Fig. 8), while in the case of CpZrCl3, one type of bridging chlorine gave rise to a narrower powder pattern than did the terminal chlorines. While further studies will be required to determine if a general relationship exists between the chlorine QI and the nature of the chlorine site, these authors were able to demonstrate, importantly, that the two different types of sites were resolvable by 35Cl SSNMR in their systems [52]. Selected chlorine quadrupolar coupling data for solid alkaline earth metal chlorides are presented in Table 4. The chlorine CS tensors have also been characterized and these data appear in Table 3. Peaks initially assigned to isotropic 35Cl chemical shifts [75] were later found to correspond to discontinuities from second-order quadrupolar powder patterns, which were present due to a nonzero QI at the chlorine nucleus [83]. The chemical shift originally reported for SrCl2, 140.8 ppm (referenced to 1 M NaCl(aq)), was found to be correct, as a result of the tetrahedral environment at

the chlorine and the resulting absence of second-order broadening due to a QI. Bryce and Bultz reported a 35/37Cl SSNMR study of several alkaline earth chlorides and their hydrates [83]. Six materials were analyzed at 11.75 and 21.1 T under static and MAS conditions, allowing for the determination of both the EFG and CS tensor parameters. An example is given in Fig. 9; the fits at multiple fields for both nuclei demonstrate the reliability of the extracted spectral parameters. Chlorine-35 CQ values were found to range from 0 MHz (SrCl2) to 4.26 ± 0.03 MHz (CaCl2  2H2O), while chlorine chemical shifts ranged from 75.0 ± 1.0 ppm (MgCl2  6H2O) to 188.2 ± 1.0 ppm (SrCl2), referenced to NaCl(s). In addition to fully characterizing the chlorine EFG and CS tensors for the materials studied, the sensitivity of chlorine SSNMR to pseudo-polymorphism was demonstrated by the significant differences in the chlorine NMR parameters of SrCl2 (CQ(35Cl)  0 MHz) and its hydrates, SrCl2  2H2O (CQ(35Cl) = 1.41 ± 0.02) and SrCl2  6H2O (CQ(35Cl) = 3.91 ± 0.05). In addition, the ability to monitor the conversion between these different pseudo-polymorphs by 35Cl SSNMR has been noted [8,83].

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Fig. 7. Chlorine-35 CT NMR spectra of n-CxH(2x+1)NH3–Cl (CxHCl) and nCxH(2x+1)ND3Cl (CxDCl) collected under stationary conditions in a magnetic field of 7.04 T. From H. Honda, Z. Naturforsch 58a (2003), 623. Reproduced by permission of the Verlag der Zeitschrift für Naturforschung.

Stebbins and co-workers collected 35Cl NMR spectra of solid powdered BaCl2 and CaCl2 during their studies of chlorine-containing glasses [84,85]. In both instances, the chlorine-35 CQ value and the chlorine chemical shift are of similar magnitude, but clearly distinct from, those observed for the hydrated salts studied by Bryce and Bultz [83]. This is another demonstration of the sensitivity of the 35Cl NMR spectrum to the hydration state of the material studied. In approximate terms, the more strongly hydrated pseudopolymorphs tend to have smaller chemical shifts. 4.3. Organic chlorides and hydrochlorides Selected chlorine quadrupolar parameters for hydrochloride salts appear in Table 5. Organic hydrochloride salts have been pop-

225

ular materials of study in the last decade, and rigorous spectral analyses have been carried out, providing both the EFG and CS tensor parameters, as demonstrated by the prevalence of these materials in Tables 3 and 5. The first hydrochloride to be analyzed by chlorine SSNMR appears to be a powdered sample of cocaine hydrochloride [86]. The 35Cl NMR spectrum was collected at 7.0 T using a Hahn-echo experiment and variable-offset data acquisition. The measured CQ(35Cl) value is 5.027 MHz. Bryce et al. later studied a series of organic hydrochlorides, and demonstrated the utility of chlorine-35/ 37 SSNMR in the study of chlorine atoms within hydrogen bonding environments [53]. Spectra for L-tyrosine HCl, L-cysteine methyl ester HCl, L-cysteine ethyl ester HCl and quinuclidine HCl were collected at 18.8 and 9.4 T, using a variety of techniques. These included Hahn-echoes under MAS and static conditions, as well as the QCPMG sequence under stationary conditions. The EFG tensor parameters were reported for all four salts, while the CS tensor parameters were given for L-cysteine ethyl ester HCl. Studies of additional amino acid hydrochlorides were reported in subsequent publications. In 2005, Gervais et al. studied four amino acid hydrochlorides using multinuclear magnetic resonance experiments, including some performed on 35Cl [87]. The authors provided the isotropic CS and EFG tensor parameters for glycine HCl, L-valine HCl and L-glutamic acid HCl and confirmed the reported parameters for L-tyrosine HCl. First principles calculations predicted relatively small chlorine CS tensor spans ranging from 78 to 157 ppm [87]. Bryce and co-workers examined the majority of the amino acid hydrochlorides in three reports published in 2006 and 2007, using these compounds as simple models for chloride ion binding sites in ion channels [48–50] (see Figs. 10 and 11). The experiments were carried out at fields of 11.75 and 21.1 T, under stationary and, when possible, MAS conditions. In addition to the EFG tensor parameters, the CS tensor parameters for the majority of these salts were determined. The extraction of CSA information benefited greatly from the ultrahigh-field experiments, as the broadening due to the QI decreases with increased field while the effects of the CSA increase. It was found that both the quadrupolar and CS tensor parameters were highly dependent on the specific local and electronic environment, as a wide range of parameters were measured, despite some similarities in the chlorine environments between different amino acid hydrochlorides. The value of CQ(35Cl) was found to generally decrease as the number of hydro-

Fig. 8. Solid-state 35Cl QCPMG NMR spectra and simulations for Cp*ZrCl3. The final simulation represents four sites (two bridging and two terminal); representative simulations of bridging and terminal sites are shown above (B0 = 9.4 T). From Rossini et al. J. Am. Chem. Soc. 131 (2009), 3317. Reproduced by permission of the American Chemical Society.

226

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Fig. 9. 35/37Cl NMR spectra of solid powdered SrCl22H2O. (a), (f) 35Cl spectra of a MAS sample obtained at 21.1 and 11.75 T, respectively. (b) and (g) show best-fit simulated spectra. (k) 37Cl spectrum of a MAS sample obtained at 11.75 T. (l) The best-fit simulated spectrum. (c), (h) 35Cl spectra of a stationary sample obtained at 21.1 and 11.75 T, respectively. The corresponding best-fit simulated spectra are shown in (d) and (i). The span of the CS tensor is set to zero in (e) and (j) to demonstrate the large effect that the anisotropy of the CS tensor has on the spectrum. (m) 37Cl spectrum of a stationary powdered sample obtained at 11.75 T. The best-fit simulated spectrum is shown in (n), and a simulated spectrum for which the CS tensor span is set to zero is shown in (o). From Bryce and Bultz, Chem. Eur. J. 13 (2007), 4786. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

gen bonds around chlorine increased; however, the symmetry around the chlorine was also an important consideration. The information available from the quadrupolar parameters and CS tensors is complementary, as there was no general correlation found between the two sets of data; for example, a larger value of CQ does not necessarily correspond to a larger CS span. The chlorine-35 CQ magnitudes range from 7.1 ± 0.1 MHz (aspartic acid HCl) to 2.03 ± 0.02 MHz (DL-arginine HCl monohydrate) [48–50]. The chlorine chemical shifts for these salts were found to be 90 to 106 ppm (referenced to NaCl(s)), with only proline HCl and serine HCl lying outside this range [48–50]. Quantum chemical calculations of the EFG and CS tensors were validated on the set of experimental data for the amino acid hydrochlorides, and then applied to estimate the chlorine CQ for the ClC chloride ion channel (Fig. 12). The value obtained, CQ(35Cl) = –2.07 MHz, indicates that future 35Cl SSNMR studies on chloride ion channels, or more sophisticated models for these channels, will not be hampered by broad CT resonances. Hamaed et al. [88] used 35Cl SSNMR to study the biologically important hydrochloride salts of four local anesthetic pharmaceuticals: procaine HCl, tetracaine HCl, monohydrated lidocaine HCl (LH), and monohydrated bupivacaine HCl (BH). Complete EFG and CS tensor parameters were provided for all materials studied, and again the sensitivity of the parameters to slight changes in structure, was observed. The chlorine-35 CQ values observed in this study range from 2.52 to 6.00 MHz while the chlorine chemical shifts range from 71 to 118 ppm (referenced to NaCl(s)), which is similar to other organic hydrochlorides [48–50,87]. Quantum chemical calculations were performed in order to further the relationship between the chlorine NMR parameters and the number of short hydrogen bonds around the chloride ion [88]. The authors demonstrated that the sensitivity of 35Cl SSNMR is great enough to easily distinguish between polymorphs of the same material (polymorphic structures of both LH and BH). Chlorine-35 NMR spectra were shown to be superior to 13C NMR and powder X-ray diffraction methods for distinguishing between polymorphs in certain cases, most notably for a pair of high-temperature polymorphs

of BH. The 35Cl NMR spectra of three BH polymorphs are shown in Fig. 13. In the context of the results of Hamaed et al., one possible application of high-field 35Cl SSNMR could be the identification of polymorphs or solvates in a pharmaceutical sample. The success of the approach will be limited by the (a priori unknown) size of the QI in a hypothetical impurity component (e.g., another polymorph, etc.), since a large QI would lead to a very broad CT signal, which would be buried in the baseline for very low concentrations, while an impurity with a very small QI would be easily detected at low concentrations due to the narrow line width of the CT in such a sample. Frydman and co-workers have introduced a two-dimensional ‘relaxation-assisted’ spectroscopic method (RAS) to resolve multiple powder patterns arising from half-integer spin quadrupolar nuclei under stationary conditions [89]. Shown in Fig. 14 is an application of the method to resolve the 35Cl resonances in a mixture of two hydrochloride salts. This method could prove to be useful in resolving multiple magnetically non-equivalent sites in a given sample; however, it relies on a significant difference in relaxation times to achieve resolution. Gordon, Brouwer, and Ripmeester have reported a 35Cl SSNMR study of several chlorine-containing ionic liquids at 9.4 and 21.1 T [90]. The four samples studied are solids at room temperature: methyl-N-butylpyridinium chloride ([MBPyri]Cl), butyldimethylimidazolium chloride ([BMMIM]Cl), ethylmethylimidazolium chloride ([EMIM]Cl), and butylmethylimidazolium chloride ([BMIM]Cl). Values of CQ(35Cl) were found to be quite small, ranging from 0.805 ± 0.005 MHz for site 2 of [EMIM]Cl to 1.500 ± 0.002 MHz for [BMIM]Cl. Chemical shift tensor spans were extracted in two cases: 77.7 ± 0.2 ppm for [BMMIM]Cl and 43.5 ± 0.5 ppm for [BMIM]Cl and are consistent with the spans observed for other organic hydrochlorides (Table 3). The 35Cl MAS NMR spectrum of [EMIM]Cl obtained at 21.1 T is impressive in that all four crystallographic sites are resolved (Fig. 15). The small values of CQ(35Cl) suggest that 79/81Br and 127I SSNMR of the bromide and iodide analogues of these ionic liquids should be feasible, assuming that the structures are isomorphic. Indeed, Gordon

227

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237 Table 5 Selected chlorine-35 quadrupolar parameters and isotropic chemical shifts for organic chlorides and hydrochlorides.a CQ(35Cl)/MHz

gQ

References

71.60(0.02) 71.65(0.05)b site 1: 91.72(0.03)b site 2: 74.98(0.04)b site 3: 71.36(0.04)b site 4: 60.60(0.03)b site 1: 83.15(0.07)b site 2: 70.56(0.04)b 94.7(0.5)c 95(1)c 117(1)c 101(5)c 114(1)c 90(10)c 104(1)c 102(1)c 9.7(10.0)d 53.2(0.5)d

0.978(0.004) 1.500(0.002) 0.808(0.004) 0.805(0.005) 0.884(0.004) 0.972(0.005) 0.857(0.008) 0.889(0.008) 2.23(0.02) 2.3(0.1) 6.5(0.1) 6.42(0.05) 6.0(0.1) 5.89(0.05) 3.7(0.1) 3.61(0.01) 5.25(0.02) 3.78(0.02)

0.10(0.02) 0.390(0.005) 0.95(0.01) 0.20(0.02) 0.86(0.01) 0.80(0.01) 0.525(0.005) 0.08(0.08) 0.72(0.03) 0.7(0.1) 0.6(0.1) 0.61(0.03) 0.5(0.1) 0.51(0.05) 0.6(0.1) 0.65(0.02) 0.05(0.01) 0.03(0.03)

Gordon et al. [90] Gordon et al. [90] Gordon et al. [90]

L-cysteine

48.2(0.7)d

2.37(0.01)

0.81(0.03)

Bryce et al. [53]

L-lysine

HCl

104.2(0.5)c 0c 105(2)c

3.92(0.01) 5.027 2.49(0.01)

0.47(0.02) 0.2(0.05) 0.42(0.02)

Chapman and Bryce [50] Yesinowski et al. [86] Bryce et al. [49]

L-serine

HCl

120(30)c

3.0 (0.3)

0.8(0.2)

Bryce et al. [49]

37(5)c

4.50(0.05)

0.63(0.05)

Bryce et al. [49]

96(20)c

4.39(0.05)

0.25(0.03)

Bryce et al. [49]

96(5)c

6.08(0.05)

0.52(0.03)

Bryce et al. [49]

105.0(1.0)c

5.05(0.04)

0.86(0.03)

Bryce and Sward [48]

91.5(1.0)c

2.035(0.020)

0.98(0.02)

Bryce and Sward [48]

106(5)c 102(5)c 93(1)a 99(1)c 99(10)c 96(6)c 71(6)c 100(4)c site 1: 85(10)c site 2: 110(10)c 96(10)c site 1: 118(10)c site 2: 95(10)c 118(5)c – – – – – – – – 14.7(10.0)d –

6.4(0.1) 7.1(0.1) 4.59(0.03) 4.41(0.02) 5.4(0.1) 4.87(0.07) 6.00(0.10) 4.67(0.07) site 1: 2.52(0.12) site 2: 5.32(0.10) 3.66(0.10) site 1: 4.75(0.20) site 2: 5.85(0.20) 4.58(0.05) 1.143(0.001) (315 K) 1.401(0.001) (300 K) 1.501(0.001) (315 K) 1.303(0.002) (320 K) 1.469(0.001) (325 K) 1.375(0.001) (300 K) 1.280(0.002) (320 K) 1.452(0.002) (325 K) 4.04(0.03) 1.2–0.85 on heating from 300–480 K

0.75(0.06) 0.42(0.05) 0.46(0.02) 0.35(0.03) 0.94(0.02) 0.28(0.04) 0.27(0.04) 0.77(0.03) 0.95(0.05) 0.32(0.10) 0.72(0.08) 0.65(0.10) 0.26(0.04) 0.56(0.06) – – – – – – – – 0.62(0.02) 0

Chapman and Chapman and Chapman and Chapman and Chapman and Hamaed et al. Hamaed et al. Hamaed et al. Hamaed et al.

Compound

diso/ppm

Butyldimethylimidazolium chloride Butylmethylimidazolium chloride Ethylmethylimidazolium chloride

Butylmethylpyridinium chloride Tyrosine HCl Glycine HCl L-valine

HCl

L-glutamic

acid HCl

Quinuclidine HCl L-cysteine

ethyl ester HCl

methyl ester HCl Cysteine HCl monohydrate Cocaine HCl

L-proline

HCl

L-isoleucine

HCl

L-phenylalanine L-tryptophan

HCl

HCl

DL-arginine

HCl monohydrate Alanine HCl Aspartic acid HCl Histidine HCl monohydrate Methionine HCl Threonine HCl Procaine HCl Tetracaine HCl Lidocaine HCl Lidocaine HCl, polymorph 1 Bupivacaine HCl Bupivacaine HCl heated to 120 °C Bupivacaine HCl heated to 170 °C n-C5H11NH3Cl n-C7H15NH3Cl n-C8H17NH3Cl n-C9H19NH3Cl n-C10H21NH3Cl n-C7H15ND3Cl n-C9H19ND3Cl n-C10H21ND3Cl Tris sarcosine CaCl2 C4H9NH3Cl; C4H9ND3Cl a b c d

b

Gordon et al. [90] Bryce et al. [53] Gervais et al. [87] Gervais et al. [87] Bryce et al. [49] Gervais et al. [87] Bryce et al. [49] Gervais et al. [87] Bryce et al. [49] Bryce et al. [53] Bryce et al. [53]

Bryce Bryce Bryce Bryce Bryce [88] [88] [88] [88]

[50] [50] [50] [50] [50]

Hamaed et al. [88] Hamaed et al. [88] Hamaed et al. [88] Honda [76] Honda [76] Honda [76] Honda [76] Honda [76] Honda [76] Honda [76] Honda [76] Bryce et al. [53] Hattori et al. [143]

Errors are given in parentheses. With respect to 0.1 M NaCl in D2O. With respect to NaCl(s). With respect to infinitely dilute NaCl(aq).

et al. have reported very promising preliminary data in this regard [91]. One of the important conclusions from the chlorine work noted above is that the SSNMR parameters are comparable to the values known for other organic hydrochlorides, which do not exhibit ionic liquid properties. Further work on these compounds could have an impact on the controversy in the literature over whether ionic liquids exhibit some unique ‘‘structure” in the liquid state [90,92–94]. 4.4. Perchlorates The experimental data for perchlorates (ClO4–) are summarized in Table 6. The chlorine site symmetry is approximately Td in all the

perchlorates studied, but the small deviation from tetrahedral symmetry results in a nonzero QI at the chlorine nuclei in these salts. The chlorine CQ values are relatively small however, and they have been popular materials to study by chlorine SSNMR [70,95– 103]. The first chlorine SSNMR spectrum of a perchlorate appears to have been published by Jurga et al. in their study of some multimethylammonium perchlorates in 1986 [95]. The 35Cl CQs in these materials were found to be relatively small, as expected, with the largest being 1.120 MHz for phase III of dimethylammonium perchlorate. In 1991 and 1992, Tarasov et al. released two publications which report on three alkali metal perchlorates [98,99]. Chlorine35 SSNMR spectra were collected at 7.04 T under stationary condi-

228

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Fig. 12. Model of the chloride ion binding site in the ClC transport channel (PDB entry 1KPL) [3]. From Bryce et al. J. Am. Chem. Soc. 128 (2006), 2121. Reproduced by permission of the American Chemical Society.

Fig. 10. Chlorine-35/37 SSNMR spectra of L-phenylalanine HCl under stationary conditions. Simulated spectra are shown above the experimental spectra. (a) 35Cl at 11.75 T; (c) 37Cl at 11.75 T; (e) 35Cl at 21.1 T; (h) 37Cl at 21.1 T. The relative orientation of the EFG and CS PASs are shown at the top. From Bryce et al. J. Am. Chem. Soc. 128 (2006), 2121. Reproduced by permission of the American Chemical Society.

of thirteen different perchlorates were analyzed using static, MAS and SATRAS methods. Both the 35Cl and 37Cl isotopes were studied. Typical CT and ST spectra for LiClO4 are shown in Fig. 16. The sensitivity of the observed chlorine NMR spectrum to the hydration state of the sample was also demonstrated (anhydrous LiClO4 compared with its trihydrate). Simulation of the chlorine MAS NMR spectra allowed for the determination of precise values of the chlorine chemical shift, CQ, and gQ values for all 13 salts. Included in this study were the perchlorates examined previously by Tarasov et al., and the CQ(35Cl) values determined in this study differ slightly, with values of 0.585, 0.537, and 0.441 MHz, for CsClO4, RbClO4, and KClO4, respectively [98,99,70]. The chemical shifts observed for the perchlorates range from 1029.6–1049.3 ppm (referenced to NaCl(s)); these are much greater than those observed for chlorides, due to increased paramagnetic contributions to the shielding tensor. 4.5. Glasses and sodalites

Fig. 11. Comparison between experimentally determined CQ(35Cl) for a series of amino acid hydrochlorides, on the x-axis, and the calculated values, on the y-axis. The equation for the line of best fit (solid line) is y = 1.0892x + 0.2055; R = 0.994. The line y = x is represented by the dashed line. From R.P. Chapman and D.L. Bryce, Phys. Chem. Chem. Phys. 9 (2007), 6219. Further details on the calculations are given in reference [50]. Reproduced by permission of the PCCP Owner Societies.

tions. CQ(35Cl) values of 0.63, 0.60, and 0.51 MHz were determined for CsClO4, RbClO4, and KClO4 respectively; however, the effects of CSA were not considered. A notable study of several perchlorates by Skibsted and Jakobsen was published in 1999 [70]. In this report, powdered samples

Experimental data for glasses, sodalites, and all other chlorinecontaining materials are summarized in Table 7. As mentioned above, Stebbins and co-workers have carried out multiple studies of chlorine-containing silicate and aluminosilicate glasses using MAS NMR at 14.1 and 18.8 T [84,85]. The chlorine-35 CQ values vary between 3.0 and 4.4 MHz in different silicate glasses and from 2.9 to 3.6 MHz in aluminosilicate glasses. The authors proposed a correlation between the chlorine chemical shift and the M–Cl bond distance. Trill et al. have published multiple studies on halogen-containing sodalites, the most extensive having been published in 2003 [46]. In this study, powdered samples of chloride-containing mixed halide sodalites were analyzed using 35Cl MAS and SATRAS NMR. The authors reported the chlorine chemical shifts and chlorine35 quadrupolar products (PQ = CQ(1 + gQ/3)1/2) for many different compositions of these sodalites, finding that PQ was very small and constant at 0.2 ± 0.1 MHz, regardless of the percentage of Cl in the cages. The authors also reported the first application of MQMAS to a chlorine-containing system, using the pulse sequence to distinguish between two different chlorine sites with chlorine

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

229

Fig. 13. Chlorine-35 SSNMR spectra at 21.1 and 9.4 T (bottom) and simulations (top) of monohydrated bupivacaine HCl (BH) polymorphs: (a) commercial BH; (b) BH1 obtained by heating BH to 120 °C; (c) BH2 obtained by heating BH to 170 °C. From H. Hamaed, J.M. Pawlowski, B.F.T. Cooper, R. Fu, S.H. Eichhorn and R.W. Schurko, J. Am. Chem. Soc. 130 (2008), 11056. Reproduced by permission of the American Chemical Society.

chemical shifts of –127.6 and –124.6 ppm (referenced to 1 M NaCl(aq)). 5. Bromine-79/81 experimental data Compared to chlorine, significantly less SSNMR data are available for 79/81Br, despite the higher receptivity of the latter nuclides. This is the result of the larger Q values for the 79/81Br nuclei, which often renders the collection of bromine SSNMR spectra difficult. However, the increased availability of high magnetic fields will likely lead to more bromine SSNMR studies. In addition, the improvement in accuracy, and increased availability, of quantum chemical computational software will likely also aid researchers, as calculations provide a method of predicting which materials will have CQ values small enough to be characterized in a given magnetic field. 5.1. Simple cubic bromides The simple cubic salts of bromine have been widely studied by SSNMR, due to the absence of any significant QI at the bromine nucleus. For bromine, the vast majority of published studies examine salts in which the bromine sits in a site where the surrounding EFG is nearly zero. The data for simple inorganic bromide salts are summarized in Tables 2 and 8. Many of the initial chlorine SSNMR studies included bromine and iodine as well. Interestingly, Pound carried out the first bro-

Fig. 14. Top: two-dimensional RAS spectroscopy of CT 35Cl powder patterns for a mixture of lysine HCl and ornithine HCl at 14.1 T [89]. The corresponding one-dimensional spectra for pure samples are shown below. From Lupulescu et al. J. Am. Chem. Soc. 125 (2003), 3376. Reproduced by permission of the American Chemical Society.

230

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Fig. 15. Chlorine-35 NMR spectra of powdered samples of the chloride salts of: (a) methyl-N-butylpyridinium; (b) butyldimethylimidazolium; (c) ethylmethylimidazolium; (d) butylmethylimidazolium. Dashed lines are best-fit simulated spectra. From P.G. Gordon, D.H. Brouwer, J.A. Ripmeester, J. Phys. Chem. A 112 (2008) 12527. Reproduced by permission of the American Chemical Society.

mine-79/81 SSNMR study on a single crystal of NaBr in 1950, years before Kanda studied the 35Cl nucleus [71,104]. Shortly thereafter, Pound and Watkins carried out experiments on KBr single crystals,

concluding that the EFG surrounding the bromine nucleus was actually not exactly zero, as a result of lattice imperfections [105,106]. The authors attributed these imperfections to lattice

Table 6 Selected chlorine-35 quadrupolar parameters and isotropic chemical shifts for perchlorates.a Compound

diso/ppm b

Cd(ClO4)2  6H2O (CH3)4NClO4 Trimethylammonnium perchlorate

1044.3(0.5) 1039.9(0.3)b 1040.1(0.5)c 917.5(0.7)d – 930d 1034.2(0.5)b 1034.1(0.5)c 1045.9(0.5)b – 1049.2(0.3)b 3(5)e 1049.4(0.3)b; 1049.1(0.3)c – 1047.7(0.3)b 1036.2(0.5)b site 1: 1046.6(0.3)b 1046.6(0.3)c site 2: 1045.5(0.3)b 1045.5(0.3)c 1029.6(0.5)b 1040.6(0.3)b 1040.5(0.5)c 1044.4(0.3)b 1049.3(0.3)b –

Dimethylammonnium perchlorate

Phase I: 1003f

Monomethylammonium perchlorate

–

NaClO4 NaClO4  H2O NH4ClO4

LiClO4 LiClO4  3H2O KClO4 RbClO4

CsClO4 Mg(ClO4)2 Mg(ClO4)2  6H2O

Ba(ClO4)2 Ba(ClO4)2  3H2O

a b c d e f

Errors are given in parentheses. With respect to NaCl(s). Refers to 37Cl value. With respect to NH4Cl(s). With respect to 0.1 M RbClO4(aq). With respect to KCl(s).

CQ(35Cl)/MHz

gQ

References

0.887(0.014) 0.566(0.009) 0.459(0.012)c 0.6949(0.0005) 0.640(0.040) (300 K) – 1.282(0.008) 1.010(0.012)c 0.695(0.004) 0.51(0.01) (296 K) 0.440(0.006) 0.6 0.537(0.015) 0.424(0.014)c 0.63 0.585(0.008) 2.981(0.007) site 1: 0.309(0.006) 0.245(0.005)c site 2: 0.475(0.008) 0.375(0.003)c 2.256(0.008) 0.383(0.005) 0.299(0.004)c 0.328(0.005) 0.307(0.004) Phase III: 370 kHz Phase II: 318 kHz Phase III: 1.120 Phase II: 0.238 Phase III: 1.016 Phase II: 0.258

0.92(0.02) 0.90(0.02) 0.91(0.04)c 0.7552(0.0012) 0.80(0.10) – 0.34(0.01) 0.34(0.01)c 0.00(0.03) 0.52(0.10) 0.88(0.02) 0.53(0.02) 0.87(0.03) 0.86(0.02)c 0.55 0.86(0.02) 0.57(0.01) 0.00(0.08) 0.00(0.10)c 0.00(0.05) 0.00(0.07)c 0.58(0.01) 0.00(0.03) 0.01(0.03)c 0.00(0.03) 0.00(0.03) Phase III: 0.60–1.0

Skibsted and Jakobsen [70] Skibsted and Jakobsen [70]

Phase Phase Phase Phase

III: 0 II: 0 III: 0.75 II: 0

Bastow and Stuart [97] Segel et al. [96] Bastow et al. [97] Skibsted and Jakobsen [70] Skibsted and Jakobsen [70] Tarasov et al. [99] Skibsted and Jakobsen [70] Tarasov et al. [99] Skibsted and Jakobsen [70] Tarasov et al. [98] Skibsted and Jakobsen [70] Skibsted and Jakobsen [70] Skibsted and Jakobsen [70]

Skibsted and Jakobsen [70] Skibsted and Jakobsen [70] Skibsted and Jakobsen [70] Skibsted and Jakobsen [70] Jurga et al. [95] Jurga et al. [95] Jurga et al. [95]

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

231

ature measurement in NMR samples over the range 20–320 K [119], echoing an earlier suggestion of using the 81Br signal of KBr to measure inhomogeneous temperature distributions in MRI experiments [120]. 5.2. Perbromates The experimental data resulting from bromine-79/81 SSNMR studies of perbromates are given in Table 9. Analogously to the perchlorates, in these perbromates the bromine sits at a position of relatively high symmetry. To the best of our knowledge, there are three studies of these materials, all carried out by Tarasov et al. [121] In these studies, four perbromate powders were examined: KBrO4, RbBrO4, CsBrO4, and NH4BrO4. All four materials were studied under static conditions (see Fig. 17 for an example). Quadrupolar coupling constants and asymmetry parameters were reported; the largest CQ(81Br) observed was 3.35 ± 0.03 MHz for KBrO4. The authors did not consider CSA in their fits. The symmetry of the systems suggests that CSA will be small, but nonzero. It would be interesting to systematically re-examine a series of perbromates in much stronger magnetic fields, in order to obtain some information on bromine CSA and how it relates to structure. 5.3. Other materials

Fig. 16. (a) 35Cl MAS NMR spectrum at 14.1 T of the CT for LiClO4 (mr = 8.0 kHz) and (c) LiClO4  3H2O (mr = 7.0 kHz). (b) Simulation of the second-order quadrupolar line shape for LiClO4. (d) 35Cl MAS NMR spectrum at 14.1 T of the STs for LiClO4  3H2O (mr = 7.0 kHz) shown with the CT cut off at approximately 1/8 of its total height. (e) Simulation of the ssb manifold in (d), employing the optimized NMR parameters for LiClO4  3H2O. From Skibsted and Jakobsen, Inorg. Chem. 38 (1999), 1806. Reproduced by permission of the American Chemical Society.

dislocations and the resulting QI ‘smeared’ the satellite transitions over a large spectral width, resulting in a decrease in the signal intensity. Kanda, Yamagata, and Hayashi and Hayamizu all included bromine NMR in their studies of simple halide salts and noted the same trends that were observed for the chlorides [71,72,74]. Many of the experiments mentioned for the chlorine cubic salts, such as examining the temperature and pressure dependence of T1 and the effect of elastic strain on the CT, were also carried out for 79/81Br [81,107–111]. In addition, there are many studies which monitored the effects of doping on the CT bromine NMR signal (as well as the diffusion of the impurities) of cubic bromide salts, which yielded the expected result that dopants decrease the intensity and increase the line width of the bromine CT [112–116]. An additional study looked at dislocation effects and how they influenced bromine self-diffusion in a freshly prepared sample of AgBr [117]. The 79Br NMR spectrum of KBr powder collected under MAS conditions features a large manifold of spinning sidebands due to the 79/81Br STs. This manifold was found to be extremely sensitive to slight changes in the magic angle setting. For this reason, it was proposed by Frye and Maciel that the 79Br NMR spectrum of powdered KBr be used to set the magic angle [118]. As bromine-79 is actually quite receptive, and the cubic symmetry of KBr results in very minimal quadrupolar broadening, a reasonable spectrum can be collected in a single scan. In addition, the proximity of the resonance frequency of 13C to that of 79Br makes the sample a convenient choice with regards to probe tuning. The 79Br chemical shift and T1 in KBr have also been proposed to be used for temper-

The experimental 79/81Br SSNMR data available for other bromine-containing systems are summarized in Table 10. Unlike the hydrochlorides, the hydrobromides have not been widely studied using SSNMR, with only three examples in the literature to the best of our knowledge. The two most recent examples were reported in a 2007 paper by Persons and Harbison, in which they used samples of leucine HBr and tyrosine HBr to demonstrate their STREAQI method, as described in Section 3.1 (Fig. 18) [63]. Bromine-79 CQ values of 49.0 and 11.26 MHz were reported for L-leucine HBr and L-tyrosine HBr, respectively. Bromine CSA was not included quantitatively in the analysis; however, the authors observed the qualitative effects of CSA, particularly in the spectrum of L-tyrosine HBr. Two alkaline earth metal bromide type systems have been studied by 81Br SSNMR: Tris–sarcosine CaBr2 [122] and BaBr2 [123]. A single crystal of tris-sarcosine CaBr2 was analyzed by Erge et al. in 1991 and a CQ(81Br) value of 21.9 MHz (gQ = 0.64) was determined. These values may be compared to CQ(35Cl) = 4.04 MHz and gQ = 0.62 for Tris–sarcosine CaCl2 [53]. On the basis of the product of the quadrupole moment and Sternheimer antishielding factor for the two nuclei, it is possible to predict that the ratio between CQ(81Br) and CQ(35Cl) should be 6.11 if the two compounds are isostructural. This is slightly higher than the experimental value of 5.42. The remaining discrepancy is likely attributable to the different unit cell sizes, as demonstrated by Wu and Terskikh for the alkali metal NMR parameters in a series of alkali metal tetraphenylborates [28]. Trill et al. also included several bromine-containing sodalites in their SSNMR studies of mixed halogen sodalite systems [124]. The bromine-81 quadrupolar products, PQ, were found to range from 0.3 to 0.7 MHz for the mixed Cl–Br sodalites, and from 0.3 to 1.0 MHz for mixed I–Br samples. Similar to the conclusion reached for chlorine, a linear correlation between d(81Br) and the M–Br bond distance was noted.

6. Iodine-127 experimental data There are comparatively few 127I SSNMR data in the literature, due in large part to its large Q (Table 1). Although it is likely that increased field strengths will lead to an increase in the number

232

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Table 7 Selected chlorine-35 quadrupolar parameters and isotropic chemical shifts for sodalites, glasses, and other materials.a Compound Mixed Cl/Br sodalites (varied from 5% Cl to 100% Cl) Mixed Cl/I sodalites (varied from 5% Cl to 100% Cl) Several Cl-containing silicate and aluminosilicate glassese Silicates NS N2CS CNS C2NS CS Alumino-silicates NAS1 NAS4 CAS1 CAS3 CAS4 Mg3Al(OH)6ClO4  nH2O LiAl2(OH)6Cl  nH2O (‘‘LiAlCl5”; ‘‘LiAlCl6”; ‘‘LiAlCl8”) Coal A (CH3)4NCdCl3

CQ(35Cl)/MHz

diso/ppm b

gQ

References

c

–125.4(0.2) (5% Cl)–123.2(0.2) (100% Cl) –128.2(0.3)b (5% Cl)–123.2(0.3) (100% Cl)

PQ = 0.2(0.1) for all Cl percentages; PQ = 55(5) kHz for pure chloro-sodalite PQ = 0.2(0.1) for all Cl percentages from MAS

–65(5)d –67(10)d –35(15)d –20(15)d 102(22)d

3.3(0.1) 3.0(0.2) 3.2(0.4) 3.3(0.3) 4.4(0.4)

89(11)d 65(7)d 52(38)d 62(45)d 79(42)d

3.0(0.3) 2.9(0.2) 3.5(0.9) 3.6(1.0) 4.0(0.9)

1001.2d (0% RH) Spectra recorded at several relative humidities (RH) and temperatures. 8.97d (0% RH) –49.6f peak assigned to NaCl –

YBa2Cu3O6.7Cl0.2 (Cl-doped YBCO) Hydrotalcite

–

Hydrocalumite

30(5)g (283–403 K) 26(5)g (<273 K)

CH3NH3GeCl3

–

Cl-doped yttrium ceramics (YBCO-Cl)

–

Calcium chloroapatite p-dichlorobenzene Ca4.99-5.06(PO4)2.98-3.00F0.510.48Cl0.38-0.36OH0.14-0.12 (CH3)2CHNH3CuCl3

Trill et al. [46] Trill et al. [46]

0.7e

Sandland et al. [85]

– 0.9 (room humidity and temperature)

– 0

Hou and Kirkpatrick[133] Hou et al.[131,132]

– Phase I (296 K): 13.48 phase I (120 K): 13.64 phase II (115 K): site 1: 14.98 site 2: 14.02 site 3: 12.64 phase III (98 K): site 1: 15.18 site 2: 14.72 site 3: 14.68 site 4: 13.98 site 5: 13.90 site 6: 13.86 site 7: 13.52 site 8: 13.30 site 9: 13.04 2

– Phase phase phase phase phase

I (296 K): 0.16 I (120 K): 0.13 II, site 1: 0.12 II, site 2: 0.22 II, site 3: 0.27

Saito et al.[144] Mulla-Osman et al. [145–147]

1

Goren et al. [148]

1 0 – 0 0 0.9 –

Kirkpatrick et al. [149]

Kirkpatrick et al. [149]

–

Amitin et al. [152]

–

2.4 (176 K) 1.5 (233 K) 1.2 (room temp) 2.87 (283 K) 2.22 (403 K) 3.0 (<273 K) 25 (low temp); ? 0 (500 K) site 1: 1.2 site 2: 1.4 site 3: 1.9 0.8

–

– 115d

73.96 1.6

0.0712(0.0005) –

Yesinowski and Eckert [86,153] Creel et al. [134] McCubbin et al. [154]

–

site 1: 19.5 site 2: 25.17 site 3: 21.05

0.68 0.5 0.58

–

Yamada et al. [150,151]

Saito et al. [155]

a

Errors are given in parentheses. With respect to 1M NaCl(aq). c PQ = CQ(1 + gQ2/3)1/2. d With respect to 1 M NaCl(aq). e CQ was determined with gQ arbitrarily set to 0.7. Estimated mean values of CQ could therefore vary by up to 15%. See Sandland et al. [85] for explanations of labels for silicates and aluminosilicates. f With respect to KCl(s). g No chemical shift reference given. b

of iodine SSNMR reports, we expect that the studies will remain limited to systems where the iodine sits in a relatively high-sym-

metry environment. As with bromine, quantum chemical calculations will play an important role, allowing researchers to fairly

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

233

Table 8 Chlorine, bromine, and iodine quadrupolar coupling constants and isotropic chemical shifts for simple inorganic halides.a compound NH4Cl NaCl NaCl/cement KCl RbCl CsCl CuCl AgCl TlCl AgBr TlBr NH4Br Phase II Phase I Phase III CuBr CuI AgI

diso/ppm b

74.0 73.8b (37Cl) –46.1b –42.0b (37Cl) 30c 3.07b 2.8b (37Cl) 44.7b 44.8b (37Cl) 109.4b 109.6b (37Cl) –127.23(0.05)d –12.82(0.05)d –250e 169.35(0.07)f 208g 620(50)h 233(10)i 132(10)i – – –116(10)j –134.14(0.2)f 7.7(0.2)k –230.2(1.0)k

CQ(81Br)/MHz

References

–

Weeding and Veeman [73]

–

Weeding and Veeman [73]

–

Barberon et al. [156]

– –

Weeding and Veeman [73] Weeding and Veeman [73]

–

Weeding and Veeman [73]

– – – – – –

Hayashi and Hayamizu [69] Hayashi and Hayamizu [69] Guenther and Hultsh [157] Hayashi and Hayamizu [69] Jelinek et al. [158] Kanda [71]

– – 5.5(0.3) 5.656(0.01) – – – –

Itoh and Yamagata [159,160]

Jeffrey et al. [161] Günther and Hultsch [157] Hayashi and Hayamizu [69,162] Hayashi and Hayamizu [69] Hayashi and Hayamizu [69]

a Errors are given in parentheses. Precise data for alkali metal halides are provided in Table 2. Data are for 35Cl, 81Br, and 127I unless otherwise indicated. b With respect to infinitely dilute Cl-. c With respect to NaCl(aq); concentration not specified. d With respect to KCl(s). e With respect to saturated NaCl(aq). f With respect to KBr(s). g With respect to 0.1 M NaBr(aq). h With respect to NaBr(aq). i With respect to a dilute aqueous alkali bromide solution. j With respect to saturated NaBr(aq). k With respect to KI(s).

Table 9 Bromine-81 quadrupolar parameters and isotropic chemical shifts for solid perbromates.a Compound

diso/ppmb

CQ(81Br)/MHz

gQ

References

KBrO4 RbBrO4 CsBrO4

– – 2400(200)

0.71(0.05) 0.37 0

Tarasov et al. [163] Tarasov et al. [163] Tarasov et al. [121]

NH4BrO4

2440(120) 2500(130)c,d –

3.35(0.03) 2.36(0.04) 1.32(0.04) 1.53(0.06)d 2.27(0.05) 2.72(0.05)d 2.38(0.03)

0.99

Tarasov et al. [164]

0.91

Tarasov et al. [163]

a b c d

Errors are given in parentheses. With respect to 1 M KBr(aq). Shift reference not given. Data for 79Br.

accurately predict whether an iodine SSNMR experiment is feasible at a given field strength. 6.1. Simple cubic iodides Selected experimental data for simple iodides are summarized in Tables 2 and 8. The high symmetry in these materials has allowed them to be relatively widely studied. The first iodine SSNMR study appears to have been carried out on a single crystal of KI [104]. In addition, 127I SSNMR experiments on KI yielded the same result that was noted in the potassium bromide case – namely that

Fig. 17. Bromine-81 CT NMR spectra for powdered KBrO4 collected under static conditions (T = 297 K, m0 = 81.256 MHz). Bottom: experimental; top: simulation. From V.P. Tarasov, V.I. Privalov, K.S. Gavrichev, V.E. Gorbunov, Yu. K. Gusev, Yu. A. Buslaev, Russ. J. Coord. Chem. 16 (1990), 854. Copyright Plenum Press. Reproduced with kind permission of Springer Science and Business Media.

the EFG about the iodine site was not perfectly zero due to lattice vibrations/dislocations [105,106]. Simple iodides were included in some of the SSNMR studies of cubic halogens described previously, such as the reports of Kanda, Yamagata, and Hayashi and Hayamizu [71,72,74]. In the MAS study of Hayashi and Hayamizu, the iodine-127 chemical shift trend was only partially consistent with the results for bromine and chlorine. Their results found the shift trend to be CsI > LiI > RbI > NaI > KI, which is inconsistent with trends in ionic radius or cation electronegativity [74]. In a detailed multinuclear magnetic resonance study of a series of silver halides, AgX (X = F, Cl, Br, and I), the authors used 19F, 35Cl, 81 Br, 127I, and 109Ag SSNMR experiments to comment upon the temperature dependence of the observed chemical shifts [125]. It was demonstrated that the overlapping ion theory, then commonly used to model observed shifts, failed to predict the anion chemical shifts of the quadrupolar halogens, while the silver shifts were in good agreement. 6.2. Periodates Along with the simple iodides, the periodates make up the majority of iodine SSNMR data available in the literature; the experimental data available for these systems are presented in Table 11. The first of these studies was published by Segel and Vyas in 1980 and reported 127I CQ values for four periodates: NH4IO4, KIO4, AgIO4, and NaIO4 [126]. The CQ values were found to range from 10.02 MHz for NH4IO4 to 42.39 MHz for NaIO4. The CS tensors were also examined in this study with some surprisingly large spans reported, the highest being 59,000 ppm. No NMR spectra, however, were shown in this publication. In 2001, Wu and Dong revisited the 127I SSNMR spectroscopy of selected periodates (HIO4, NH4IO4, KIO4, AgIO4, and NaIO4) [127]. The authors noted that the very large CS tensor spans reported by Segel and Vyas [126] for NaIO4 seemed suspect due to the symmetry in the system, and then demonstrated that the CSA was actually less than 50 ppm for all five periodates. An exact

234

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

Table 10 Selected bromine-81 quadrupolar parameters and isotropic chemical shifts for sodalites, organic hydrobromides, and other materials.a Compound Mixed Cl/Br sodalites (varied from 10% to Br to 100% Br) Mixed Br/I sodalites (varied from 6% to Br to 100% Br) Na8[Al6Si6O24]Brn(H3O2)2n (varied from n = 0.980.10) Na6+n[Al6Si6O24]Brn(4H2O)2n (varied from n = 0.980.10) Na6+n[Al6Si6O24]Brnu2n (u = empty sodalite cage) (varied from n = 0.980.10) Na8[Al6Si6O12]Br2 Na8npAgn[Si6Al6O12]Br2p (various n and y) Ag8[Al6Si6O12]Br2 K2Pt(CN)4Br0.33.2 H2O I II BaFBr:O (2%) Cn–DABCO–Cn–Br2 (n = 10,12, and 14) (DABCO = 1,4-diazoniabicyclo[2.2.2]octane) Tris–sarcosine CaBr2 BaBr2 C60Br24Br2 Deuterated glycyl-L-alanine HBr  H2O L-leucine

HBr

L-tyrosine a b c d e f

HBr

CQ(81Br)/MHz

diso/ppm b

gQ

References

–220.6(0.4) to –217.3(0.4) –226.4(0.4)b to –220.6(0.4) –221.0b to –218.0 –221.0b to – 230.6

PQc

= 0.7(0.2)–0.3(0.2) PQ = 0.3(0.2)–1.0(0.2) PQ = 0.5–0.8 PQ = 0.5–2.2

Trill Trill Trill Trill

–221.0b to –237.8

PQ = 0.5–1.0

Trill et al. [124]

-219f –219/550f –550f

– – –

– – –

Jelinek et al. [158,165] Jelinek et al. [158,165] Jelinek et al. [158,165]

– – – 3660d,e

17.6 15.1 10.00 –

0 0 0 –

Brenni et al. [166] Brinkmann et al. [167] Bastow et al. [141] Nakayama et al. [168]

– – – – –

21.9 – – 19.750 49.0e

0.64 – – 0.8328 0.59

Erge et al. [122] Potrepka et al. [123] Milia et al. [169,170] Kehrer et al. [171] Persons and Harbison [63]

–

11.26e

0.86

Persons and Harbison [63]

et et et et

al. [46] al.[46] al.[124] al. [124]

Errors are given in parentheses. With respect to 1 M NaBr(aq). PQ = CQ(1 + gQ2/3)1/2. With respect to NaBr(s). Data for 79Br. With respect to 0.1 M NaBr(aq).

Fig. 18. 79Br STREAQI NMR spectra of L-tyrosine HBr at 14 T, simulated using the assumption (a) of a discrete jump of the rotor between two static orientations and (b) continuous sample rotation; compared with (c) an experimental spectrum. From J. Persons and G.S. Harbison, J. Magn. Reson. 186 (2007) 347. Reproduced by permission of Elsevier Limited.

value of 18 ppm was determined for CsIO4 [127]. This remains as what is likely the only reliable iodine CS tensor span determined by SSNMR spectroscopy. The 127I CQ in several periodates was shown to be highly dependent on temperature in a series of studies by Burkert and co-workers in the 1980s [128]. For example, in the case of NH4IO4, CQ was found to vary from 1.31 to 11.59 MHz over the temperature range 145– 440 K [128]. A consistent relationship was not found amongst different periodates however, as the CQ value increased with increasing temperature in some cases, while decreasing in others [128].

at 0.936 T, polycrystalline IF6+AsF6– was examined by 127I SSNMR, which was feasible due to the fairly symmetric environment about the iodine nucleus in the cation [129]. The authors were able to determine that the iodine-127 CQ value was between 2.32 and 2.9 MHz. In addition, Trill et al. have also used 127I SSNMR to study several mixed halogen sodalite systems at 11.7 T [46]. The authors determined that the 127I quadrupolar products in the mixed Cl–I sodalite systems varied from 1.0 to 3.9 MHz and from 1.0 to 2.8 MHz in the mixed Br–I sodalites. 7. Outlook and conclusions

6.3. Other materials The available iodine SSNMR data for other materials are summarized in Table 11. Few materials other than simple iodides and periodates have been studied. However, in a 1970 study done

Significant advances in the application of 35/37Cl solid-state NMR spectroscopy, and the interpretation of the results in terms of local structure, have occurred in the past few years. The majority of recent studies have understandably focused on chloride ions in

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

235

Table 11 Iodine-127 quadrupolar parameters and isotropic chemical shifts for periodates and other materials.a compound KIO4 NH4IO4

RbIO4 NaIO4 CsIO4

HIO4 (CH3)4AsIO4 (n-C4H9)4PIO4 (C2H5)4SbIO4 (C2H5)4PIO4 (C2H5)4AsIO4 (C6H5)4PIO4 (C6H5)4AsIO4 (C6H5)4SbIO4 (CH3)4NIO4 (n-C4H9)4NIO4 N(CH3)4IO4 AgIO4 Mixed Br/I sodalites (varied from 0% Br–90% Br) Mixed Cl/I sodalites (varied from 0% Cl–90% Cl) IF7 IF6+AsF6– Mn(CO)5I a b c d e f

diso/ppm b

3960(10) 6530(200)c 3960(10)b  3810(50)c 3960(10)b  3950(10)b –5370(100)c 3972(2)b  3300(10)b  – – – – – – – – – – 5410(50)c –255.5(0.8)d to –245.0(0.8) –255.5(0.8)d to –227.1(0.8) 3040(40)f – –

CQ(127I)/MHz

gQ

references

20.66(0.01) 20.73 10.00(0.01) 1.31–11.59 (145–440 K) 10.02 15.65(0.01) 16.27–14.65 (202–433 K) 42.24(0.01) 42.39 1.00(0.01) 11.45–10.45 (202–294 K); 1.8–1.7 (300–315 K) 43.00(0.01) 61.8 5.98–3.14 (150–355 K) 5.29–5.71 (217–301 K) 6.09–5.87 (177–302 K) 5.82–5.55 (197–299 K) 4.12–5.71 (228–293 K) 3.50–7.42 (179–297 K) 1.37–2.76 (228–302 K) 20–15 (200–300 K) 8.07–3.35 (225–310 K) 15.46–9.07 (301–418 K) 29.66 PQe = 1.0(0.5)–2.8(0.5) PQ = 1.0(0.5)–4.3(0.5) 134.88(0.03) (77 K) 2.9–2.32 927

0  0   0  0  0 

Wu and Dong [127] Segel and Vyas [126] Wu and Dong [127] Burkert [128] Segel and Vyas [126] Wu and Dong [127] Burkert [128] Wu and Dong [127] Segel and Vyas [126] Wu and Dong [127] Burkert [128]

0.75  0.08 to 0.30 0.64 to 0.38 – – 0 0 0 – 0.60 to 0.77 – –

Wu and Dong [127] Grommelt and Burkert [172] Burkert and Grommelt [32] Klobasa and Burkert [173] Klobasa et al. [174] Klobasa et al. [174] Burkert and Klobasa; Klobasa et al. [174,175] Burkert and Klobasa; Klobasa et al. [174,175] Burkert and Klobasa; Klobasa et al. [174,175] Klobasa and Burkert [173] Burkert and Grommelt [32] Burkert [128] Segel and Vyas [126] Trill et al. [46] Trill et al. [46] Weulersse et al. [176] Hon and Christe [129] Spiess and Sheline [137]

0.04(0.004) – 0.03

Errors are given in parentheses. With respect to KI(s). With respect to NaI(s). With respect to 1 M KI(aq). PQ = CQ(1 + gQ2/3)1/2. With respect to 5 M KI(aq).

various organic and inorganic environments, due to the relatively small QI of the chloride nucleus. This article has highlighted what we believe to be the most important recent developments, e.g., the interpretation of 35/37Cl quadrupolar and CS tensors in organic hydrochlorides in terms of hydrogen bonding environment; the demonstration of the sensitivity of 35/37Cl SSNMR to polymorphism and pseudopolymorphism in materials ranging from pharmaceutical hydrochlorides to inorganic chlorides; the characterization of solid ionic ‘liquids’ via chlorine SSNMR. Covalently bound chlorine presents more of a challenge for modern SSNMR methods, particularly for chlorine bound to carbon, where the value of CQ(35Cl) typically exceeds 70 MHz [13,82]. However, when chlorine is covalently bound to transition metals, it has been shown that the observed CQ(35Cl) are much smaller [52]. Overall, it is seen that ultrahigh magnetic field strengths (P18.8 T) have been extremely beneficial to the above-mentioned studies. By and large, it is clear that the 35/37Cl nuclides will remain the most accessible to study by SSNMR, and it is for these nuclides that the most comprehensive understanding of the relationship between local structure and the observed NMR spectrum is available. We anticipate continued advances in this area, with further applications to a range of systems comprising organic, biomolecular, inorganic, and organometallic compounds. New developments in the field of 79/81Br and 127I SSNMR spectroscopy have been scarce due to the large quadrupole moments and Sternheimer antishielding factors associated with these nuclei. Nevertheless, for the bromine nuclides in particular, ultrahigh magnetic fields should create new opportunities for study. Variable-offset approaches will likely be required to make progress. Potential alternatives include the STREAQI and WURST approaches [58,59,63]. Regardless of the method used, there is still work to

be done to build a better understanding of the structural and electronic factors which influence the quadrupolar and shielding interactions for 79/81Br and 127I. For example, there is essentially no information concerning bromine CS tensors, with a report by Fusaro and Doane being the lone exception [130] in solids, and only one reliable report for 127I [127]. Therefore, while the spectroscopy will remain challenging due to large second-order quadrupolar broadening for all but the most symmetric of environments, there is also a great deal of room for development and exploration in this area of NMR spectroscopy. Acknowledgements D.L.B. thanks the Natural Sciences and Engineering Research Council (NSERC) of Canada for funding. C.M.W. thanks NSERC for an Alexander Graham Bell CGS D2 scholarship. We are grateful to Aaron Rossini and Prof. Rob Schurko for providing a preprint of reference [52]. We thank Prof. Gang Wu for helpful discussions. Access to the 900 MHz NMR spectrometer (Fig. 6) was provided by the National Ultrahigh-Field NMR Facility for Solids (Ottawa, Canada), a national research facility funded by the Canada Foundation for Innovation, the Ontario Innovation Trust, Recherche Québec, the National Research Council Canada, and Bruker BioSpin and managed by the University of Ottawa (www.nmr900.ca). NSERC is acknowledged for a Major Resources Support grant. References [1] F.M. Ashcroft, Ion Channels and Disease, Academic Press, San Diego, 2000. [2] M.E. Loewen, G.W. Forsyth, Physiol. Rev. 85 (2005) 1061. [3] R. Dutzler, E.B. Campbell, M. Cadene, B.T. Chait, R. MacKinnon, Nature 415 (2002) 287.

236

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237

[4] R. Dutzler, E.B. Campbell, R. MacKinnon, Science 300 (2003) 108. [5] T.M. Trnka, J.P. Morgan, M.S. Sanford, T.E. Wilhelm, M. Scholl, T.-L. Choi, S. Ding, M.W. Day, R.H. Grubbs, J. Am. Chem. Soc. 125 (2002) 2546. [6] K. Mikami, M. Terada, H. Matsuzawa, Angew. Chem. Int. Ed. 41 (2002) 3554. [7] D.L. Bryce, G.D. Sward, Magn. Reson. Chem. 44 (2006) 409. [8] C.M. Widdifield, R.P. Chapman, D.L. Bryce, Chlorine, bromine, and iodine solid-state NMR spectroscopy, in: G.A. Webb (Ed.), Annual Reports on Nuclear Magnetic Resonance Spectroscopy, vol. 66, Elsevier Ltd, 2009, pp. 195–326. [9] B. Lindman, S. Forsén, Chlorine, bromine and iodine NMR. Physico-chemical and biological applications, in: P. Diehl, E. Fluck, R. Kosfeld (Eds.), NMR Basic Principles and Progress, vol. 12, Springer-Verlag, Berlin, 1976, pp. 1–360. [10] B. Lindman, S. Forsén, The halogens – chlorine, bromine and iodine, in: R.K. Harris, B.E. Mann (Eds.), NMR and the Periodic Table, Academic Press, London, 1978, pp. 421–436. [11] T. Drakenberg, S. Forsén, NATO ASI Ser. Ser. C 103 (1983) 405. [12] J. Mason, Multinuclear NMR, Plenum Press, New York, 1987. [13] E.A.C. Lucken, Nuclear Quadrupole Coupling Constants, Academic Press, London, 1969. [14] A. Abragam, Principles of Nuclear Magnetism, Oxford University Press, New York, 1961. [15] C.P. Slichter, Principles of Magnetic Resonance, Springer-Verlag, New York, 1990. [16] K. Schmidt-Rohr, H.W. Spiess, Multidimensional Solid-State NMR and Polymers, Academic Press, San Diego, 1994. [17] K.J.D. MacKenzie, M.E. Smith, Multinuclear Solid-State NMR of Inorganic Materials, Pergamon, Amsterdam, 2002. [18] M.J. Duer, Solid-State NMR Spectroscopy, Blackwell Publishing, Oxford, 2004. [19] M.H. Cohen, F. Reif, Solid State Phys. 5 (1957) 321. [20] U. Haeberlen, Advances in Magnetic Resonance, Supplement 1, Academic Press, New York, 1976. [21] J. Mason, Solid State Nucl. Magn. Reson. 2 (1993) 285. [22] H.W. Spiess, Rotation of molecules and nuclear spin relaxation, in: P. Diehl, E. Fluck, R. Kosfeld (Eds.), NMR Basic Principles and Progress, vol. 15, SpringerVelag, Berlin, 1978, pp. 55–214. [23] M. Mehring, Principles of High Resolution NMR in Solids, Springer-Verlag, New York, 1983. [24] J. Herzfeld, A.E. Berger, J. Chem. Phys. 73 (1980) 6021. [25] G.B. Arfken, Mathematical Methods for Physicists, Academic Press, New York, 1985. [26] R. Sternheimer, Phys. Rev. 80 (1950) 102. [27] R. Sternheimer, Phys. Rev. 84 (1951) 244. [28] G. Wu, V. Terskikh, J. Phys. Chem. A 112 (2008) 10359. [29] A.D. Bain, beta software, private communication, 2008. [30] R.K. Harris, E.D. Becker, S.M. Cabral de Menezes, R. Goodfellow, P. Granger, Pure Appl. Chem. 73 (2001) 1795. [31] J. Blaser, O. Lutz, W. Steinkilberg, Z. Naturforsch. A 27 (1972) 72. [32] P.K. Burkert, M. Grommelt, Z. Naturforsch. 43b (1988) 1381. [33] A. Loewnstein, M. Shporer, P.C. Lauterbur, J.E. Ramirez, J. Chem. Soc. Chem. Commun. (1968) 214. [34] M. Gee, R.E. Wasylishen, A. Laaksonen, J. Phys. Chem. A 103 (1999) 10805. [35] C.Y. Lee, C.D. Cornwell, Magnetic Resonance and Related Phenomena, Groupement Ampere, Heidelberg, 1976. p. 261. [36] W. Gauß, S. Gunther, A.R. Haase, M. Kerber, D. Kessler, J. Kronenbitter, H. Kruger, O. Lutz, A. Nolle, P. Schrade, M. Schule, G.E. Siegloch, Z. Naturforsch. A 33 (1978) 934. [37] J. Kondo, J. Yamashita, J. Phys. Chem. Solids 10 (1959) 245. [38] J. Itoh, Y. Yamagata, J. Phys. Soc. Jpn. 14 (1959) 225. [39] R. Baron, J. Chem. Phys. 38 (1963) 173. [40] D. Ikenberry, T.P. Das, Phys. Rev. 138 (1965) A822. [41] J. Itoh, Y. Yamagata, J. Phys. Soc. Jpn. 13 (1958) 1182. [42] D.L. Bryce, R.E. Wasylishen, Acc. Chem. Res. 36 (2003) 327. [43] M.A.M. Forgeron, R.E. Wasylishen, G.H. Penner, J. Phys. Chem. A 108 (2004) 4751. [44] L. Frydman, J.S. Harwood, J. Am. Chem. Soc. 117 (1995) 5367. [45] Z. Gan, J. Am. Chem. Soc. 122 (2000) 3242. [46] H. Trill, H. Eckert, V.I. Srdanov, J. Phys. Chem. B 107 (2003) 8779. [47] F.H. Larsen, H.J. Jakobsen, P.D. Ellis, N.C. Nielsen, J. Phys. Chem. A 101 (1997) 8597. [48] D.L. Bryce, G.D. Sward, J. Phys. Chem. B. 110 (2006) 26461. [49] D.L. Bryce, G.D. Sward, S. Adiga, J. Am. Chem. Soc. 128 (2006) 2121. [50] R.P. Chapman, D.L. Bryce, Phys. Chem. Chem. Phys. 9 (2007) 6219. [51] C.M. Widdifield, D.L. Bryce, (unpublished work). [52] A.J. Rossini, R.W. Mills, G.A. Briscoe, E.L. Norton, S.J. Geier, I. Hung, S. Zheng, J. Autschbach, R.W. Schurko, J. Am. Chem. Soc. 131 (2009) 3317. [53] D.L. Bryce, M. Gee, R.E. Wasylishen, J. Phys. Chem. A 105 (2001) 10413. [54] R. Siegel, T.T. Nakashima, R.E. Wasylishen, Concept Magn. Reson. A 26 (2005) 62. [55] R. Siegel, T.T. Nakashima, R.E. Wasylishen, Concept Magn. Reson. A 26 (2005) 47. [56] A.P.M. Kentgens, R. Verhagen, Chem. Phys. Lett. 300 (1999) 435. [57] R. Siegel, T.T. Nakashima, R.E. Wasylishen, Chem. Phys. Lett. 388 (2004) 441. [58] L.A. O’Dell, R.W. Schurko, Chem. Phys. Lett. 464 (2008) 97. [59] L.A. O’Dell, A.J. Rossini, R.W. Schurko, Chem. Phys. Lett. 468 (2009) 330. [60] A. Samoson, Chem. Phys. Lett. 119 (1985) 29.

[61] C. Jäger, Satellite transition spectroscopy of quadrupolar nuclei, in: P. Diehl, E. Fluck, H. Günther, R. Kosfeld, J. Seelig (Eds.), NMR Basic Principles and Progress, vol. 31, Springer-Verlag, Berlin, 1994, pp. 133–170. [62] J. Skibsted, N.C. Nielsen, H. Bildsøe, H.J. Jakobsen, J. Magn. Reson. 95 (1991) 88. [63] J. Persons, G.S. Harbison, J. Magn. Reson. 186 (2007) 347. [64] K. Eichele, R.E. Wasylishen, WSOLIDS NMR Simulation Package 1.17.30, 2001. [65] D.W. Alderman, M.S. Solum, D.M. Grant, J. Chem. Phys. 84 (1986) 3717. [66] M. Bak, J.T. Rasmussen, N.C. Nielsen, J. Magn. Reson. 147 (2000) 296. [67] D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calvé, B. Alonso, J.-O. Durand, B. Bujoli, Z. Gan, G. Hoatson, Magn. Reson. Chem. 40 (2002) 70. [68] P. Pyykkö, Mol. Phys. 106 (2008) 1965. [69] S. Hayashi, K. Hayamizu, J. Phys. Chem. Solids 53 (1992) 239. [70] J. Skibsted, H.J. Jakobsen, Inorg. Chem. 38 (1999) 1806. [71] T. Kanda, J. Phys. Soc. Jpn. 10 (1955) 85. [72] Y. Yamagata, J. Phys. Soc. Jpn. 19 (1964) 10. [73] T.L. Weeding, W.S. Veeman, J. Chem. Soc. Chem. Commun. (1989) 946. [74] S. Hayashi, K. Hayamizu, Bull. Chem. Soc. Jpn. 63 (1990) 913. [75] F. Lefebvre, J. Chim. Phys. 89 (1992) 1767. [76] H. Honda, Z. Naturforsch. 58a (2003) 623. [77] T. Yamanishi, T. Kanashiro, A. Itabashi, Y. Michihiro, Y. Kishimoto, T. Ohno, J. Phys. Soc. Jpn. 63 (1994) 3903. [78] T. Yamanishi, T. Kanashiro, Y. Michihiro, Y. Kishimoto, T. Ohno, J. Phys. Soc. Jpn. 64 (1995) 643. [79] Y. Michihiro, T. Yamanishi, T. Kanashiro, Y. Kishimoto, Solid State Ionics 79 (1995) 40. [80] T. Kanashiro, Y. Michihiro, K. Kitahara, T. Yamanishi, Y. Kishimoto, T. Ohno, Solid State Ionics 86–88 (1996) 223. [81] J.L. Marsh Jr., P.A. Casabella, Phys. Rev. 150 (1966) 546. [82] K. Eichele, R.E. Wasylishen, J.S. Grossert, A.C. Olivieri, J. Phys. Chem. 99 (1995) 10110. [83] D.L. Bryce, E.B. Bultz, Chem. Eur. J. 13 (2007) 4786. [84] J.F. Stebbins, L.S. Du, Am. Mineral. 87 (2002) 359. [85] T.O. Sandland, L.S. Du, J.F. Stebbins, J.D. Webster, Geochem. Cosmochim. Acta 68 (2004) 5059. [86] J.P. Yesinowski, M.L. Buess, A.N. Garroway, M. Ziegeweld, A. Pines, Anal. Chem. 67 (1995) 2256. [87] C. Gervais, R. Dupree, K.J. Pike, C. Bonhomme, M. Profeta, C.J. Pickard, F. Mauri, J. Phys. Chem. 109 (2005) 6960. [88] H. Hamaed, J.M. Pawlowski, B.F.T. Cooper, R. Fu, S.H. Eichhorn, R.W. Schurko, J. Am. Chem. Soc. 130 (2008) 11056. [89] A. Lupulescu, M. Kotecha, L. Frydman, J. Am. Chem. Soc. 125 (2003) 3376. [90] P.G. Gordon, D.H. Brouwer, J.A. Ripmeester, J. Phys. Chem. A 112 (2008) 12527. [91] P.G. Gordon, D.H. Brouwer, J.A. Ripmeester, Probing the local structure of ionic liquid salts with 35Cl, 79Br and 127I solid state NMR. 40th Inorganic Discussion Weekend, University of Toronto, Toronto, 2007, p. 21. [92] C. Hardacre, S.E.J. Holbrey, J.D. McMath, D.T. Bowron, A.K. Soper, J. Chem. Phys. 118 (2003) 273. [93] S. Tsuzuki, H. Tokuda, M. Mikami, Phys. Chem. Chem. Phys. 9 (2007) 4780. [94] H. Katayanagi, S. Hayashi, H. Hamaguchi, K. Nishikawa, Chem. Phys. Lett. 392 (2004) 460. [95] S. Jurga, J. Seliger, R. Blinc, H.W. Spiess, Phys. Lett. A 116 (1986) 295. [96] S.L. Segel, S. Maxwell, R.D. Heyding, P. Ingman, E. Ylinen, M. Punkkinen, Solid State Commun. 66 (1988) 1039. [97] T.J. Bastow, S.N. Stuart, J. Phys. Condens. Matter 1 (1989) 4649. [98] V.P. Tarasov, M.A. Meladze, G.A. Kirakosyan, A.E. Shvelashvili, Y.A. Buslaev, Phys. Status Solidi B 167 (1991) 271. [99] V.P. Tarasov, M.A. Meladze, G.A. Kirakosyan, Russ. J. Coord. Chem. 18 (1992) 823. [100] N. Binesh, S.V. Bhat, Solid State Ionics 92 (1996) 261. [101] H. Ono, S. Ishimaru, R. Ikeda, H. Ishida, Chem. Phys. Lett. 275 (1997) 485. [102] H. Ono, S. Ishimaru, R. Ikeda, H. Ishida, Bull. Chem. Soc. Jpn. 72 (1999) 2049. [103] O. Czupin´ski, M. Wojtas´, J. Zaleski, R. Jakubas, W. Medycki, J. Phys. Condens. Matter 18 (2006) 3307. [104] R.V. Pound, Phys. Rev. 79 (1950) 685. [105] G.D. Watkins, R.V. Pound, Phys. Rev. 89 (1953) 658. [106] R.V. Pound, J. Phys. Chem. 57 (1953) 743. [107] M.J. Weber, Phys. Rev. 130 (1963) 1. [108] R.R. Allen, M.J. Weber, J. Chem. Phys. 38 (1963) 2970. [109] M.J. Weber, R.R. Allen, Proc. Coll. AMPERE 11 (1963) 177. [110] B.D. Günther, M.L. Simon, R.A. Hultsch, B. Am. Phys. Soc. Ser. 2 (14) (1969) 188. [111] T.C. Ng, E. Tward, J. Magn. Reson. 51 (1983) 361. [112] F. Reif, Phys. Rev. 100 (1955) 1597. [113] E. Otsuka, H. Kawamura, J. Phys. Soc. Jpn. 12 (1957) 1071. [114] M.E. Melich, W.D. Ohlsen, B. Am. Phys. Soc. Ser. 2 (8) (1963) 468. [115] W.D. Ohlsen, M.E. Melich, Phys. Rev. 144 (1966) 240. [116] K.D. Becker, J. Phys.-Paris, Colloque 41 (1980) C6–249. [117] O.A.P. Tavares, B. D’Aguanno, L.G. Conti, Gazz. Chim. Ital. 113 (1983) 529. [118] J.S. Frye, G.E. Maciel, J. Magn. Reson. 48 (1982) 125. [119] K.R. Thurber, R. Tycko, J. Magn. Reson. 196 (2009) 84. [120] B.H. Suits, D. White, J. Appl. Phys. 60 (1986) 3772. [121] V.P. Tarasov, S.A. Petrushin, Y.K. Gusev, Dokl. Akad. Nauk SSSR 293 (1987) 1423.

R.P. Chapman et al. / Progress in Nuclear Magnetic Resonance Spectroscopy 55 (2009) 215–237 [122] T. Erge, D. Michel, J. Petersson, F. Engelke, Phys. Status Solidi A 123 (1991) 325. [123] D.M. Potrepka, J.I. Budnick, D.B. Fenner, W.A. Hines, M. Balasubramanian, A.R. Moodenbaugh, Phys. Rev. B 60 (1999) 10489. [124] H. Trill, H. Eckert, V.I. Srdanov, J. Am. Chem. Soc. 124 (2002) 8361. [125] K.D. Becker, E. von Goldammer, Chem. Phys. 48 (1980) 193. [126] S.L. Segel, H.M. Vyas, J. Chem. Phys. 72 (1980) 1406. [127] G. Wu, S. Dong, Solid State Nucl. Magn. Reson. 20 (2001) 100. [128] P.K. Burkert, Z. Naturforsch. 35b (1980) 1349. [129] J.F. Hon, K.O. Christe, J. Chem. Phys. 52 (1970) 1960. [130] R. Fusaro, J.W. Doane, J. Chem. Phys. 47 (1967) 5446. [131] X. Hou, R.J. Kirkpatrick, Inorg. Chem. 40 (2001) 6397. [132] X. Hou, A.G. Kalinichev, R.J. Kirkpatrick, Chem. Mater. 14 (2002) 2078. [133] X. Hou, R.J. Kirkpatrick, Chem. Mater. 14 (2002) 1195. [134] R.B. Creel, E. von Meerwall, C.F. Griffin, R.G. Barnes, J. Chem. Phys. 58 (1973) 4930. [135] A. Kawamori, J. Itoh, J. Phys. Soc. Jpn. 18 (1963) 1614. [136] F. Tabak, A. Lascialfari, A. Rigamonti, J. Phys Conden. Matter 5 (1993) B31. [137] H.W. Spiess, R.K. Sheline, J. Chem. Phys. 54 (1971) 1099. [138] R.G. Barnes, S.L. Segel, J. Chem. Phys. 37 (1962) 1895. [139] W.H. Jones, S.L. Segel, Phys. Rev. Lett. 13 (1964) 528. [140] S.H. Choh, C.V. Stager, Can. J. Phys. 49 (1971) 144. [141] T.J. Bastow, S.N. Stuart, W.G. McDugle, R.S. Eachus, J.M. Spaeth, J. Phys. Condens. Matter 6 (1994) 8633. [142] C. Brinkmann, S. Faske, M. Vogel, T. Nilges, A. Heuer, H. Eckert, Phys. Chem. Chem. Phys. 8 (2006) 369. [143] M. Hattori, Y. Onoda, T. Erata, M.E. Smith, M. Hattori, H. Ohki, R. Ikeda, Z. Naturforsch. 49a (1994) 291. [144] K. Saito, K. Kanehashi, I. Komaki, Annu. Rep. Nucl. Magn. Reson. Spectrosc. 44 (2001) 23. [145] S. Mulla-Osman, D. Michel, Z. Czapla, W.D. Hoffman, J. Phys. Condens. Matter 10 (1998) 2465. [146] S. Mulla-Osman, D. Michel, G. Völkel, Z. Czapla, Phys. Status Solidi B 219 (2000) 9. [147] S. Mulla-Osman, D. Michel, G. Völkel, I. Peral, G. Madariaga, J. Phys. Condens. Matter 13 (2001) 1119. [148] S.D. Goren, L.F. Ben-Yakar, A. Shames, B. Pandyopadhayay, C. Korn, H. Shaked, P. Massiot, C. Perrin, J. Gallier, A. Privalov, Physica C 313 (1999) 127. [149] R.J. Kirkpatrick, P. Yu, X. Hou, Y. Kim, Am. Mineral. 84 (1999) 1186. [150] K. Yamada, K. Isobe, T. Okuda, Y. Furukawa, Z. Naturforsch. 49a (1994) 258. [151] K. Yamada, K. Isobe, E. Tsuyama, T. Okuda, Y. Furukawa, Solid State Ionics 79 (1995) 152.

237

[152] E.B. Amitin, N.V. Bausck, S.A. Gromilov, S.G. Kozlova, N.K. Moroz, L.N. Mazalov, V.N. Naumov, P.P. Samoilov, S.A. Slobodjan, M.A. Starikov, V.E. Fedorov, G.I. Frolova, S.B. Erenburg, Physica C 209 (1993) 407. [153] J.P. Yesinowski, H. Eckert, Unpublished data, 1985. [154] F.M. McCubbin, H.E. Mason, H. Park, B.L. Phillips, J.B. Parise, H. Nekvasil, D.H. Lindsley, Am. Mineral. 93 (2008) 210. [155] T. Saito, H. Inoue, J. Tonisi, A. Oosawa, T. Goto, T. Sasaki, N. Kobayasi, S. Awaji, K. Watanabe, J. Phys. Conf. Ser. 51 (2006) 203. [156] F. Barberon, V. Baroghel-Bouny, H. Zanni, B. Bresson, J.B. d’Espinose de la Caillerie, L. Malosse, Z. Gan, Magn. Reson. Imaging 23 (2005) 267. [157] B.D. Günther, R.A. Hultsch, J. Magn. Reson. 1 (1969) 609. [158] R. Jelinek, A. Stein, G.A. Ozin, J. Am. Chem. Soc. 115 (1993) 2390. [159] J. Itoh, Y. Yamagata, J. Chem. Phys. 24 (1956) 621. [160] J. Itoh, Y. Yamagata, J. Phys. Soc. Jpn. 17 (1962) 481. [161] K.R. Jeffrey, A.G. Brown, R.L. Armstrong, Phys. Rev. B 8 (1973) 3071. [162] S. Hayashi, K. Hayamizu, J. Chem. Phys. 92 (1990) 2818. [163] V.P. Tarasov, V.I. Privalov, K.S. Gavrichev, V.E. Gorbunov, Y.K. Gusev, Y.A. Buslaev, Russ. J. Coord. Chem. 16 (1990) 854. [164] V.P. Tarasov, S.A. Petrushin, Y.K. Gusev, Russ. J. Inorg. Chem. 33 (1988) 452. [165] R. Jelinek, B.F. Chmelka, A. Stein, G.A. Ozin, Multinuclear solid-state NMR study of a sodalite semiconductor supralattice. Mater. Res. Soc. Symp. Proc., San Francisco, California, USA, 1992, p. 235. [166] P. Brenni, D. Brinkmann, H. Huber, M. Mali, J. Roos, H. Arend, Solid State Commun. 47 (1983) 415. [167] D. Brinkmann, H. Huber, M. Mali, J. Roos, H. Arend, Helv. Phys. Acta 55 (1982) 568. [168] H. Nakayama, T. Eguchi, N. Nakamura, H. Chihara, T. Nogami, K. Imamura, Y. Shirota, Bull. Chem. Soc. Jpn. 62 (1989) 399. [169] F. Milia, G. Papavassiliou, A. Leventis, G. Papantopoulos, P. Venturini, D. Mihailovic, J. Pirnat, R. Blinc, Br NMR study of C60Br24Br2. Proceedings of the International Winterschool on Electronic Properties of Novel Materials, Kirchber, Austria, 1995, pp. 470–473. [170] F. Milia, G. Papavassiliou, A. Leventis, J. Dolinsek, 13C and 81Br NMR in C60Br24Br2. Proceedings of the International Winterschool on Electronic Properties of Novel Materials, Kirchberg, Austria, 1995, p. 159. [171] A. Kehrer, N. Weiden, A. Weiss, Z. Phys. Chem. 178 (1992) 1. [172] M. Grommelt, P.K. Burkert, Z. Naturforsch. 44b (1989) 1053. [173] D.G. Klobasa, P.K. Burkert, Z. Naturforsch. 42a (1987) 275. [174] D.G. Klobasa, P.K. Burkert, G. Müller, Z. Naturforsch. 41a (1986) 330. [175] P.K. Burkert, D.G. Klobasa, Z. Naturforsch. 40a (1985) 274. [176] J.M. Weulersse, J. Virlet, P. Rigny, Mol. Phys. 38 (1979) 923. [177] R.K. Harris, E.D. Becker, S.M. Cabral De Menezes, P. Granger, R.E. Hoffman, K.W. Zilm, Pure Appl. Chem. 80 (2008) 59.